into the core DLL, since it's not much code and lots of stuff uses
it (e.g., try getting a tempfile name without it).

Guido has in mind an easier way for users to code this stuff, but the
only tests we have now are for fixed-offset tzinfo classes, and this
stuff is extremely delicate in the endcases (read the new test code
for why:  there are holes in time <wink>).

core generator for random.py.

M PyShell.py

1. PyShell Rev 1.39, EditorWindow Rev 1.37 fix was not handling a
   multiline prompt.
2. The same fix introduced a bug where hitting <enter> at a previous
   prompt-only line would copy the prompt to the iomark.
3. Move the setting of sys.ps1 earlier, into PyShell.main(), to allow
   this code to work before a shell is started up.
4. If cursor is on the input line in the prompt, and you hit <enter>,
   process the line instead of complaining.
5. If line has no stdin range (this includes the last line before shell
   restart) strip any prompt before recalling.

Also, made some whitespace cleanup.

Python 2.2.x backport candidate. (This bug has been around since
Python 1.6.)

Suggested by MvL.

equality.  Note, there is another flavor that compares to a given
number of significant digits rather than decimal places.  If there
is a demand, that could be added at a later date.

Obtain cleaner coding and a system wide
performance boost by using the fast, pre-parsed
PyArg_Unpack function instead of PyArg_ParseTuple
function which is driven by a format string.

experimental.

This can cause core dumps when Python runs.  Python relies on the 754-
(and C99-) mandated default "non-stop" mode for FP exceptions.  This
patch from Ben Laurie disables at least one FP exception on FreeBSD at
Python startup time.

operands have identical tzinfo members (meaning object identity -- "is").
I misunderstood the intent here, reading wrong conclusion into
conflicting clues.

window) to a combobox listing the known interpreters.

correct FSSpec implementations.

Carbon.File version, so the old method names work.

MethodGenerator, this should be OSErrMethodGenerator.

subtraction, work as documented.  In the Python implementation,
they weren't calling utcoffset() if both operands had the same
tzinfo object.  That's fine if it so happens that the shared
tzinfo object returns a fixed offset (independent of operand),
but can give wrong results if that's not so, and the latter
obtains in a tzinfo subclass instance trying to model both
standard and daylight times.  The C implementation was already
doing this "correctly", so we're just adding tests to verify it.

on OSX then the actual error (file not found) was obscured by the
error message that tried to be helpful about the allowed arguments. Fixed.

normal 2.3 framework install) and MacPython-OSX-$(VERSION) (for the
experimental Jaguar addon install).

module.

The code is shorter, more readable, faster, and dramatically increases the
range of acceptable dates.

Also, used the floor division operator in leapdays().