date:20110521

Re: Overuse of try/except/else?

2011-05-21 Thread James Mills

On Sat, May 21, 2011 at 3:40 PM, Cameron Simpson  wrote:
> These days I think I'd use a LateFunction (a facility of my own which is
> a lot like the futures module) which returns a callable when you submit
> a function; the worker thread runs the submitted function and catches the
> return value or raised exception. Anyone who calls the returned callable
> later gets the return value or the exception reraised as appropriate,
> so one can avoid the dangerous "catch everything and log" scenario.

I like this :) I guess you call this a Proxy Object although
your calling it a LateFunction seems more "fitting" :)

cheers
James

-- 
-- James Mills
--
-- "Problems are solved by method"
-- 
http://mail.python.org/mailman/listinfo/python-list

Re: TK program problem

2011-05-21 Thread Peter Otten

bvdp wrote:

> I've just done an update to my system here to Ubuntu 11.04. Mostly no
> problems ... but I have an important (to me) python/TK program that's
> stopped working. Well, it works ... mostly.
> 
> The python version is 2.7.1+ (no idea what the + means!).
> 
> I _think_ I have traced the problem to certain menus which call a
> class. The calls appear to be ignored.
> 
> Basically, what I have is a line like:
> 
> bf = makeButtonBar(root, row=0, column=0, buttons=(
>  ("Quit", self.quitall ),
>  ("Stop", self.stopPmidi ),
>  ("New Dir", self.chd),
>  ("Load Playlist", self.playList),
>  ("Favorites", selectFav),
>  ("Options", setOptions) ) )
> 
> To create a menu bar. The function makeButtonBar() creates the buttons
> with:
> 
> for txt, cmd in buttons:
> Button(bf, text=txt, height=1, command=cmd).grid(column=c,
> row=0, pady=5)
> 
> 
> All this is fine (and worked perfectly before my upgrade). The menu
> items which are ordinary functions continue to work. BUT the callbacks
> which are classes are just ignored when they are clicked.
> 
> A cut from one of the ignored classes:
> 
> 
> class selectFav:
> 
> def __init__(self):
> ...
> 
> And I've inserted some prints in the __init__() and nothing is
> printed. Also, converted the class to new-style () but no change there
> either.
> 
> Either python/tk has changed or my system is totally $*(#*#.
> Suggestions welcome!

Here's a minimal script to reproduces the problem:

$ cat tkcallclass.py
import Tkinter as tk

root = tk.Tk()
root.withdraw()

class Classic:
def __init__(self):
print "hello"

button = tk.Button(root, command=Classic)
button.invoke()
$ python2.6 tkcallclass.py
hello
$ python2.7 tkcallclass.py
Traceback (most recent call last):
  File "tkcallclass.py", line 11, in 
button.invoke()
  File "/usr/local/lib/python2.7/lib-tk/Tkinter.py", line 2081, in invoke
return self.tk.call(self._w, 'invoke')
_tkinter.TclError: invalid command name "__main__.Classic"
$

In 2.7 the Tkinter code was changed to use hasattr(obj, "__call__") instead 
of callable(obj) to recognize callbacks. This gives different results for 
oldstyle classes

>>> class A: pass
...
>>> callable(A)
True
>>> hasattr(A, "__call__")
False

...and they are no longer registered automatically with Tkinter. In theory 
you could register them explicitly yourself

$ cat tkcallclass2.py
import Tkinter as tk

root = tk.Tk()
root.withdraw()

class Classic:
def __init__(self):
print "hello"

button = tk.Button(root, command=root.register(Classic))
button.invoke()
$ python2.7 tkcallclass2.py
hello

but in practice changing them to newstyle (i. e. have them inherit from 
object) or wrapping them in a lambda appears convenient.

Personally, I would reconsider whether using a class as a callback is really 
necessary. Replacing the class with a function should be a straightforward 
process.

-- 
http://mail.python.org/mailman/listinfo/python-list

Re: hash values and equality

2011-05-21 Thread Gregory Ewing


Ethan Furman wrote:

Ulrich Eckhardt wrote:

If two equal objects have different hashes, they 
will be stored in different places in the hash map. Looking for 
object1 will then not turn up with object2, even though they are equal. 


In this case this is the behavior I want.


You can't rely on it, though. The hash value gets reduced
modulo the size of the dict, so even if two objects have
different hashes, in some cases they will land on the same
dict slot anyway.

So an object such as you're postulating would behave
unpredictably when used as a dict key. Sometimes a lookup
using a different but equal object would find it, and
sometimes not, seemingly at random.

--
Greg
--
http://mail.python.org/mailman/listinfo/python-list

Python in CS1

2011-05-21 Thread Franck Ditter

Except at MIT, who knows some good CS1 references for teaching Python ?
Thanks,

   franck
-- 
http://mail.python.org/mailman/listinfo/python-list

distutils on windows ignoring package_data

2011-05-21 Thread Wilbert Berendsen

Hi,

I have created a distutils setup.py script that installs a list of packages 
(all subpackages of one main package). Some packages have package_data files 
(icons, translations, etc).

Running setup.py install on my Linux system (python 2.6) perfectly installs 
all package data, but on Windows (tested with both 2.6 and 2.7) the 
package_data is not installed at all. Also not picked up in build, or 
bdist_wininst, etc. Only the Python files in the packages are installed.

The setup.py file can be seen at
https://github.com/wbsoft/frescobaldi/blob/master/setup.py

I later added the package_dir incantation, but it doesn't help on Windows and 
on Linux it worked perfectly without.

Does anybody have a clue why Python distutils on Windows ignore the 
package_data?

(Note: sdist perfectly works as the data files are also listed in MANIFEST.in)
tia,
Wilbert Berendsen

-- 
http://www.wilbertberendsen.nl/
-- 
http://mail.python.org/mailman/listinfo/python-list

Re: Faster Recursive Fibonacci Numbers

2011-05-21 Thread Gregory Ewing


Chris Angelico wrote:

It seems
strange to smoothly slide from native integer to long integer and just
keep on going, and yet to be unable to do the same if there's a
fractional part on it.


The trouble is that if you always compute exact results
by default, the number of digits required can blow up
very quickly.

There's a story that ABC (the predecessor to Python)
calculated everything using rationals. People found that
sometimes a calculation would take an unexpectedly long
time, and it turned out to be because internally it was
creating fractions with huge numerators and denominators.

As a consequence, Guido decided that Python would *not*
use rationals by default.

The same problem doesn't arise with ints, because the
only time you get an int with a large number of digits
is when you genuinely need a large int. So expanding
ints automatically to however many digits is needed
doesn't do any harm.

In Python 2.6 and later, there's a fractions module for
when you need it.

--
Greg
--
http://mail.python.org/mailman/listinfo/python-list

Re: English Idiom in Unix: Directory Recursively

2011-05-21 Thread Jonathan de Boyne Pollard

The supposed inefficiency of recursive implementations is based
largely on the properties of hardware that is now obsolete. With
modern processors there's no great efficiency hit. In some of the
smaller microcontrollers, it's true, you do have to worry about stack
overflow; but the ARM processors, for example, provide plenty of stack
space.

In the microcontroller world, the big performance hits come from the
fact that the only available compilers are for C and sometimes C++.
(And nobody uses assembly language except for the very little jobs.)
The nature of the C language prevents compilers from doing
optimisations that are standard in compilers for high-level languages.
Most C compilers will, for example, always pass parameters on the
stack, despite the generous supply of registers available in newer
hardware.

However, some C compilers will *also* have one or more "go faster"
calling conventions that pass parameters in registers, which can be
employed. Over in the PC world, such "go faster" calling conventions
are even the default calling convention if no calling convention is
explicitly specified, for some C and C++ compilers.

67 matches

Mail list logo