The doc string for cStringIO suggested that str() of a StringIO object
was equivalent to getvalue(). This was never true, so repair the doc
string. (doctest would have helped here.)
Bug fix candidate for any past versions.
mmap_find_method(): this obtained the string to find via s#, but it
ignored its length, acting as if it were \0-terminated instead.
Someone please run on Linux too (the extended test_mmap works on Windows).
Bugfix candidate.
over SEP, ALTSEP and MAXPATHLEN.
Patched up posixmodule.c for MSVC, but unsure what the story is now on
other non-Unixish platforms -- the preprocessor maze has no exit <wink>.
Modules/
posixmodule.c
- use SEP,ALTSEP #defines instead of hard coded path separator chars
- use EMX specific variants of chdir2(),getcwd() that support drive letters
- OS/2+EMX spawnv(),spawnve() support
- EMX specific popen[234]() derived from Win32 popen[234]() code
socketmodule.c. No code outside of the .c file references it, so it
doesn't belong the .h file (at least not yet ...), and declaring it
an imported symbol in the .h file can't be made to work on Windows (it's
a cross-DLL symbol then) without substantial code rewriting. Also
repaired the comment that goes along with the decl, to stop referring
to names and functions that haven't existed for 7 years <wink>.
socketmodule.c compiles cleanly on Windows again. The test_socket dies
at once, though (later).
helper module _ssl.
The support for the RAND_* APIs in _ssl is now only enabled
for OpenSSL 0.9.5 and up since they were added in that
release.
Note that socketmodule.* should really be renamed to _socket.* --
unfortunately, this seems to lose the CVS history of the file.
Please review and test... I was only able to test the header file
chaos in socketmodule.c/h on Linux. The test run through fine
and compiles don't give errors or warnings.
WARNING: This patch does *not* include changes to the various
non-Unix build process files.
where their capabilities intersect. Would be nice if people using non-
MSVC compilers (Borland etc) took a whack at doing something similar for
them (this code relies on the MS _cwait function).
[ #510644 ] test_curses segfaults
If we use the *object* *allocator*, we should use the *object* *deallocator*,
not the *raw memory* deallocator (confused yet?).
I think this was what caused segfaults when pymalloc was enabled.
Even if it wasn't the cause, it's still wrong.
2.2.1 candidate.
[ #504284 ] Last build problems on AIX
I'm ignoring the suggestion that this should be an autoconf test in the
interests of having a fix today. Feel free to quibble.
new.instancemethod() -- the instancemethod object is now a perfectly
general container.
This fixes SF bug ##503091 (Pedro Rodriquez): new.instancemethod fails
for new classes
This is a 2.2.1 candidate.
Patch from Mark Hammond, plus code rearrangement and comments from me.
posix_do_stat(): Windows-specific code could try to free() stack
memory in some cases when a path ending with a forward or backward slash
was passed to os.stat().
metaclass, reported by Dan Parisien.
Objects that are instances of custom metaclasses, i.e. whose ob_type
is a subclass of PyType_Type, should be pickled the same as new-style
classes (objects whose ob_type is PyType_Type). This can't be done
through the existing dispatch switches, and the __reduce__ trick
doesn't work for these, since it finds the unbound __reduce__ for
instances of the class (inherited from PyBaseObject_Type). So check
explicitly using PyType_IsSubtype().
binascii_b2a_base64(): We didn't allocate enough buffer space for very
short inputs (e.g., a 1-byte input can produce a 5-byte output, but we
only allocated 2 bytes). I expect that malloc overheads absorbed the
overrun in practice, but computing a correct upper bound is a very simple
change.
got a barrage of compile errors that didn't make sense to the C++ brain:
MSVC does not allow C (but does allow C++) initializers to contain
data addresses supplied by other DLLs. So changed the initializers here
to use dummy nulls, and changed module init to plug in the foreign
addresses at runtime (manually simulating what C++ does by magic). Tested
on Windows, and Guido tested on Linux (thanks!). BTW, the *point* is that
people are going to use this module as a template for writing their own
subtypes, and it's unusual for extension authors to build their extensions
into Python directly (separate DLLs are the norm on Windows); so it's
better if we give them a template that works <wink>.
work with Mac OS X Aqua-Tk, all nicely within ifdefs.
The process is not for the faint of heart, though: you need to download
and install the (alfa) Aqua-Tk, obtain a few needed X11 headers from
somewhere else and then everything builds. To run scripts using Tkinter
you must build with --enable-framework, build Python.app in Mac/OSX
and run your Tkinter scripts with that. Then, about half the tests in
Demo/tkinter work (or at least do something).
Checking this in anyway because it shouldn't break anything, and newer
versions of Aqua-Tk will streamline the process.
backed out of broken minimal repeat patch from July
also fixed a couple of minor potential resource leaks in pattern_subx
(Guido had already fixed the big one)
type.__module__ behavior.
This adds the module name and a dot in front of the type name in every
type object initializer, except for built-in types (and those that
already had this). Note that it touches lots of Mac modules -- I have
no way to test these but the changes look right. Apologies if they're
not. This also touches the weakref docs, which contains a sample type
object initializer. It also touches the mmap test output, because the
mmap type's repr is included in that output. It touches object.h to
put the correct description in a comment.
The OSS Programmer's Reference (www.4front-tech.com)
states:
*Setting Sampling Parameters
There are three parameters which affect the sound
quality (and therefore memory and bandwidth
requirements) of sampled audio data. These are:
** sample format (sometimes called number of bits)
** number of channels (mono or stereo), and
** sampling rate (speed)
NOTE:
It is important to always set these parameters in the
above order. Setting sampling rate before the number
of channels doesn't work with all devices.
Anthony Roach.
Release the global interpreter lock around platform spawn calls.
Bugfix candidate? Hard to say; I favor "yes, bugfix".
These clearly *should* have been releasing the GIL all along, if for no
other reason than compatibility with the similar os.system(). But it's
possible some program out there is (a) multithreaded, (b) calling a spawn
function with P_WAIT, and (c) relying on the spawn call to block all their
threads until the spawned program completes. I think it's very unlikely
anyone is doing that on purpose, but someone may be doing so by accident.
casts with a variable oself that has the proper type. A smart
compiler may put this thing into a register.
(I'm not sure what good this does except satisfy my desire to
understand this function; I got a report about an uninitialized read
from Insure++ about this function and it hurt my eyes to even look at
it. I gotta run away or I'll get tempted to reformat the entire
file...)
(At least for the repeatable test case that Tim produced.)
pattern_subx(): Add missing DECREF(filter) in both exit branches
(normal and error return). Also fix a DECREF(args) that should
certainly be a DECREF(match) -- because it's inside if (!args) and
right after allocation of match.
The st_future slot of the symtable is not freed by PySymtable_Free()
because it is shared by the symtable and compiling structs in
compiel.c. Since it is shared, it is explicitly deallocated when the
compiling struct is freed.
Big Hammer to implement -Qnew as PEP 238 says it should work (a global
option affecting all instances of "/").
pydebug.h, main.c, pythonrun.c: define a private _Py_QnewFlag flag, true
iff -Qnew is passed on the command line. This should go away (as the
comments say) when true division becomes The Rule. This is
deliberately not exposed to runtime inspection or modification: it's
a one-way one-shot switch to pretend you're using Python 3.
ceval.c: when _Py_QnewFlag is set, treat BINARY_DIVIDE as
BINARY_TRUE_DIVIDE.
test_{descr, generators, zipfile}.py: fiddle so these pass under
-Qnew too. This was just a matter of s!/!//! in test_generators and
test_zipfile. test_descr was trickier, as testbinop() is passed
assumptions that "/" is the same as calling a "__div__" method; put
a temporary hack there to call "__truediv__" instead when the method
name is "__div__" and 1/2 evaluates to 0.5.
Three standard tests still fail under -Qnew (on Windows; somebody
please try the Linux tests with -Qnew too! Linux runs a whole bunch
of tests Windows doesn't):
test_augassign
test_class
test_coercion
I can't stay awake longer to stare at this (be my guest). Offhand
cures weren't obvious, nor was it even obvious that cures are possible
without major hackery.
Question: when -Qnew is in effect, should calls to __div__ magically
change into calls to __truediv__? See "major hackery" at tail end of
last paragraph <wink>.
parser_tuple2st() and a failure to propogate an error in
build_node_children() (masking yet another leak, of course!).
This closes SF bug #485133 (confirmed by Insure++).
Bugfix candidate.
A numerically naive computation of output buffer size caused crashes
and spurious MemoryErrors for reasonable arguments.
audioop_ratecv(): Avoid spurious overflow by careful reworking of the
buffer size computations, triggering MemoryError if and only if the
final buffer size can't be represented in a C int (although
PyString_FromStringAndSize may legitimately raise MemoryError even if
it does fit in a C int). All reasonable arguments should work as
intended now, and all unreasonable arguments should be cuaght.
find_class(): We no longer mask all exceptions[1] by transforming them
into SystemError. The latter is definitely not the right thing to do,
so we let any exceptions that occur in the PyObject_GetAttr() call to
simply propagate up if they occur.
[1] Note that pickle only masked ImportError, KeyError, and
AttributeError, but cPickle masked all exceptions.
This gives mmap() on Windows the ability to create read-only, write-
through and copy-on-write mmaps. A new keyword argument is introduced
because the mmap() signatures diverged between Windows and Unix, so
while they (now) both support this functionality, there wasn't a way to
spell it in a common way without introducing a new spelling gimmick.
The old spellings are still accepted, so there isn't a backward-
compatibility issue here.
objects to save in gc.garbage. This should be the last change needed to
fix SF bug 477059: "__del__ on new classes vs. GC".
Note that this change slightly changes the behavior of the collector.
Before, if a cycle was found that contained instances with __del__
methods then all instance objects in that cycle were saved in
gc.garbage. Now, only objects with __del__ methods are saved in
gc.garbage.
When moving objects with a __del__ attribute to a special list, look
for __del__ on new-style classes with the HEAPTYPE flag set as well.
(HEAPTYPE means the class was created by a class statement.)
to lists of values, giving the contents of all the ADD_INFO records
seen so far. This is initialized agressively when the log file is
opened, so that whoever is looking at the log reader can always see
the initial data loaded into the data stream. ADD_INFO events later
in the log file continue to be reported to the application layer as
before.
Add a new method, addinfo(), to the profiler. This can be used to
insert additional ADD_INFO records into the profiler log.
Fix the tp_flags and tp_name slots on the type objects.
"socket.socket" -- on Windows, "socket.socket" is the wrapper class.
Also added the module name to the SSL type (which is not a new-style
class -- I don't want to mess with it yet).
constructor acts just like socket() before. All three arguments have
a sensible default now; socket() is equivalent to
socket(AF_INET, SOCK_STREAM).
One minor issue: the socket() function and the SocketType had
different doc strings; socket.__doc__ gave the signature,
SocketType.__doc__ gave the methods. I've merged these for now, but
maybe the list of methods is no longer necessary since it can easily
be recovered through socket.__dict__.keys(). The problem with keeping
it is that the total doc string is a bit long (34 lines -- it scrolls
of a standard tty screen).
Another general issue with the socket module is that it's a big mess.
There's pages and pages of random platform #ifdefs, and the naming
conventions are totally wrong: it uses Py prefixes and CapWords for
static functions. That's a cleanup for another day... (Also I think
the big starting comment that summarizes the API can go -- it's a
repeat of the docstring.)
error occurs, and doesn't return a count. (This is my second patch
from SF patch #474307, with small change to the docstring for send().)
2.1.2 "bugfix" candidate.
Replace some tortuous code that was trying to be clever but forgot to
DECREF the key and value, by more longwinded but obviously correct
code.
(Inspired by but not copying the fix from SF patch #475033.)
STRICT_SYSV_CURSES when compiling curses module on HP/UX. Generalize
access to _flags on systems where WINDOW is opaque. Fixes bugs
#432497, #422265, and the curses parts of #467145 and #473150.
strings, not C strings)
removed USE_PYTHON defines, and related sre.py helpers
skip calling the subx helper if the template is callable.
interestingly enough, this means that
def callback(m):
return literal
result = pattern.sub(callback, string)
is much faster than
result = pattern.sub(literal, string)
removed (conceptually flawed) getliteral helper; the new sub/subn code
uses a faster code path for literal replacement strings, but doesn't
(yet) look for literal patterns.
added STATE_OFFSET macro, and use it to convert state.start/ptr to
char indexes
Allow passing strings to the .border() method
Correct some error messages ("1 or 4" -> "1 to 4")
Bump version number
Tweak code formatting
Update my e-mail address
This adds unsetenv to posix, and uses it in the __delitem__ method of
os.environ.
(XXX Should we change the preferred name for putenv to setenv, for
consistency?)
This was submitted by Moshe, but apparently he's too busy to check it
in himself. He wrote:
Here is a function in GNU readline called add_history,
which is used to manage the history list. Though Python
uses this function internally, it does not expose it to
the Python programmer. This patch adds direct interface
to this function with documentation.
This could be used by friendly modules to "seed" the
history with commands.
This is a big one, touching lots of files. Some of the platforms
aren't tested yet. Briefly, this changes the return value of the
os/posix functions stat(), fstat(), statvfs(), fstatvfs(), and the
time functions localtime(), gmtime(), and strptime() from tuples into
pseudo-sequences. When accessed as a sequence, they behave exactly as
before. But they also have attributes like st_mtime or tm_year. The
stat return value, moreover, has a few platform-specific attributes
that are not available through the sequence interface (because
everybody expects the sequence to have a fixed length, these couldn't
be added there). If your platform's struct stat doesn't define
st_blksize, st_blocks or st_rdev, they won't be accessible from Python
either.
(Still missing is a documentation update.)
Quoth the OpenSSL RAND_add man page:
OpenSSL makes sure that the PRNG state is unique for each
thread. On systems that provide /dev/urandom, the
randomness device is used to seed the PRNG transparently.
However, on all other systems, the application is
responsible for seeding the PRNG by calling RAND_add(),
RAND_egd(3) or RAND_load_file(3).
I decided to expose RAND_add() because it's general and RAND_egd()
because it's a useful special case. RAND_load_file() didn't seem to
offer much over RAND_add(), so I skipped it. Also supplied
RAND_status() which returns true if the PRNG is seeded and false if
not.
Apparently this patch (rev 2.41) replaced all the good old "s#"
formats in PyArg_ParseTuple() with "S". Then it did
PyString_FromStringAndSize() to get back the values setup by the
"s#" format. It also incref'd and decref'd the string obtained by
"S" even though the argument tuple had a reference to it.
Replace PyString_AsString() calls with PyString_AS_STRING().
A good rule of thumb -- if you never check the return value of
PyString_AsString() to see if it's NULL, you ought to be using the
macro <wink>.
Many functions used a local variable called return_error, which was
initialized to zero. If an error occurred, it was set to true. Most
of the code paths checked were only executed if return_error was
false. goto is clearer.
The code also seemed to be written under the curious assumption that
calling Py_DECREF() on a local variable would assign the variable to
NULL. As a result, more of the error-exit code paths returned an
object that had a reference count of zero instead of just returning
NULL. Fixed the code to explicitly assign NULL after the DECREF.
A bit more reformatting, but not much.
XXX Need a much better test suite for zlib, since it the current tests
don't exercise any of this broken code.
This changes Pythread_start_thread() to return the thread ID, or -1
for an error. (It's technically an incompatible API change, but I
doubt anyone calls it.)
Mostly by Toby Dickenson and Titus Brown.
Add an optional argument to a decompression object's decompress()
method. The argument specifies the maximum length of the return
value. If the uncompressed data exceeds this length, the excess data
is stored as the unconsumed_tail attribute. (Not to be confused with
unused_data, which is a separate issue.)
Difference from SF patch: Default value for unconsumed_tail is ""
rather than None. It's simpler if the attribute is always a string.
Added support for saving the names of the functions observed into the
profile log.
Added support for using the profiler to measure coverage without collecting
timing information (which is the slow part). Coverage logs can also be
substantially smaller than profiling logs where per-line information is
being collected.
Updated comments on the log format; corrected record type values in some
of the record descriptions.
Raise ValueError when an object contains an arbitrarily nested
reference to itself. (The previous fix just produced invalid
pickles.)
Solution is very much like Py_ReprEnter() and Py_ReprLeave():
fast_save_enter() and fast_save_leave() that tracks the fast_container
limit and keeps a fast_memo of objects currently being pickled.
The cost of the solution is moderately expensive for deeply nested
structures, but it still seems to be faster than normal pickling,
based on tests with deeply nested lists.
Once FAST_LIMIT is exceeded, the new code is about twice as slow as
fast-mode code that doesn't check for recursion. It's still twice as
fast as the normal pickling code. In the absence of deeply nested
structures, I couldn't measure a difference.
To whoever who changed a bunch of (PyCFunction) casts to
(PyNoArgsFunction) in PyMethodDef initializers: don't do that. The
cast is to shut the compiler up. The compiler wants the function
pointer initializer to be a PyCFunction.
"for <var> in <testlist> may no longer be a single test followed by
a comma. This solves SF bug #431886. Note that if the testlist
contains more than one test, a trailing comma is still allowed, for
maximum backward compatibility; but this example is not:
[(x, y) for x in range(10), for y in range(10)]
^
The fix involved creating a new nonterminal 'testlist_safe' whose
definition doesn't allow the trailing comma if there's only one test:
testlist_safe: test [(',' test)+ [',']]
a misunderstanding of the refcont behavior of the 'O' format code in
PyArg_ParseTuple() and Py_BuildValue(), respectively.
- pobj is only a borrowed reference, so should *not* be DECREF'ed at
the end. This was the cause of SF bug #470635.
- The Py_BuildValue() call would leak the object produced by
makesockaddr(). (I found this by eyeballing the code.)
Add a fast_container member to Picklerobject. If fast is true, then
fast_container counts the depth of nested container calls. If the
depth exceeds FAST_LIMIT (2000), the fast flag is ignored and the
normal checks occur. This approach is much like the approach for
prevent stack overflow for comparison and reprs of recursive objects
(e.g. [[...]]).
- Fast container used for save_list(), save_dict(), and
save_inst().
XXX Not clear which other save_xxx() functions should use it.
Make Picklerobject into new-style types, using PyObject_GenericGetAttr()
and PyObject_GenericSetAttr().
- Use PyMemberDef for binary and fast members
- Use PyGetSetDef for persistent_id, inst_persistent_id, memo, and
PicklingError.
XXX Not all of these seem like they need to use getset, but it's
not clear why the old getattr() and setattr() had such odd
semantics. One change is that the getvalue() attribute will
exist on all Picklers, not just list-based picklers; I think
this is a more rationale interface.
There is a long laundry list of other changes:
- Remove unused #defines for PyList_SET_ITEM() etc.
- Make some of the indentation consistent
- Replace uses of cPickle_PyMapping_HasKey() where the first
argument is self->memo with calls to PyDict_GetItem(), because
self->memo must be a dictionary.
- Don't bother to check if cPickle_PyMapping_HasKey() returns < 0,
because it can only return 0 or 1.
- Replace uses of PyObject_CallObject() with PyObject_Call(), when
we can guarantee that the argument tuple is really a tuple.
Performance impacts of these changes:
- 5% speedup for normal pickling
- No change to fast-mode pickling.
XXX Really need tests for all the features in cPickle that aren't in
pickle.
The platform requires 8-byte alignment for doubles, but the GC header
was 12 bytes and that threw off the natural alignment of the double
members of a subtype of complex. The fix puts the GC header into a
union with a double as the other member, to force no-looser-than
double alignment of GC headers. On boxes that require 8-byte alignment
for doubles, this may add pad bytes to the GC header accordingly; ditto
for platforms that *prefer* 8-byte alignment for doubles. On platforms
that don't care, it shouldn't change the memory layout (because the
size of the old GC header is certainly greater than the size of a double
on all platforms, so unioning with a double shouldn't change size or
alignment on such boxes).
Use #define X509_NAME_MAXLEN for server/issuer length on an SSL
object.
Update doc strings for socket.ssl() and ssl methods read() and
write().
PySSL_SSLwrite(): Check return value and raise exception on error.
Use int for len instead of size_t. (All the function the size_t obj
was passed to our from expected an int!)
PySSL_SSLread(): Check return value of PyArg_ParseTuple()! More
robust checks of return values from SSL_read().
Change all the local names that start with SSL to start with PySSL.
The OpenSSL library defines lots of calls that start with "SSL_". The
calls for Python's SSL objects also started with "SSL_". This choice
made it really confusing to figure out which calls were to the library
and which calls were local to the file.
Add PySSL_SetError() that sets an exception based on the information
from SSL_get_error(). This function will eventually replace all the
calls that set it with an error message that is based on the name of
the call that failed rather than the reason it failed. (Example: If
SSL_connect() failed it used to report "SSL_connect error" now it will
offer a specific message about why SSL_connect failed.)
XXX It might be helpful to augment the error message generated
below with the name of the SSL function that generated the error.
I expect it's obvious most of the time.
Remove several unnecessary INCREFs in the module's constructor call.
PyDict_SetItem() and friends do the INCREF for you.
In SSL_dealloc(), free/dealloc them only if they're non-NULL.
Fixes some obvious core dumps, but not sure yet if there are more
semantics to the SSL calls that would affect the dealloc.
XXX [1] These changes aren't tested very thoroughly, because regrtest
doesn't do any SSL tests. I've done some trivial tests on my own, but
don't really know how to use the key and cert files. In one case, an
SSL-level error causes Python to dump core. I'll get the fixed in the
next round of changes.
XXX [2] The checkin removes the x_attr member of the SSLObject struct.
I'm not sure if this is kosher for backwards compatibility at the
binary level. Perhaps its safer to keep the member but keep it
assigned to NULL.
And the leaks?
newSSLObject() called PyDict_New(), stored the result in x_attr
without checking it, and later stored NULL in x_attr without doing
anything to the dict. So the dict always leaks. There is no further
reference to x_attr, so I just removed it completely.
The error cases in newSSLObject() passed the return value of
PyString_FromString() directly to PyErr_SetObject().
PyErr_SetObject() expects a borrowed reference, so the string leaked.
This simplifies the rounding in _PyObject_VAR_SIZE, allows to restore the
pre-rounding calling sequence, and allows some nice little simplifications
in its callers. I'm still making it return a size_t, though.
As Guido suggested, this makes the new subclassing code substantially
simpler. But the mechanics of doing it w/ C macro semantics are a mess,
and _PyObject_VAR_SIZE has a new calling sequence now.
Question: The PyObject_NEW_VAR macro appears to be part of the public API.
Regardless of what it expands to, the notion that it has to round up the
memory it allocates is new, and extensions containing the old
PyObject_NEW_VAR macro expansion (which was embedded in the
PyObject_NEW_VAR expansion) won't do this rounding. But the rounding
isn't actually *needed* except for new-style instances with dict pointers
after a variable-length blob of embedded data. So my guess is that we do
not need to bump the API version for this (as the rounding isn't needed
for anything an extension can do unless it's recompiled anyway). What's
your guess?
Jeremy Hylton
Tim Peters
Martin v. Löwis
Andrew MacIntyre
Guido van Rossum
Marc-André Lemburg
Fred Drake
Michael W. Hudson
Mark Hammond
Jack Jansen
Neal Norwitz
Barry Warsaw
Fredrik Lundh
Andrew M. Kuchling
Neil Schemenauer