🌐 Mailman 3 March 2005 - Python-list - python.org

Re: UnicodeEncodeError in string conversion
by Serge Orlov March 31, 2005

March 31, 2005

Maurice LING wrote: > Hi, > > I'm working on a script that searches a public database and retrives > results using SOAP interface. However, it seems that the results may > contains unicodes. When I try to pump the results into Firebird database > using kinterbasdb module, some results will give me a > UnicodeEncodeError. From what I've gathered, it looks like the > conversion from the results from SOAP interface to string results in the > error. > > Is there any way to get thru this? Have you read this FAQ entry: http://kinterbasdb.sourceforge.net/dist_docs/usage.html#faq_fep_unicode ?? Serge.

1 0

More decorator rumination
by Scott David Daniels March 31, 2005

March 31, 2005

Over on comp.python.education we were discussing a (generally seen as) misuse of decorators to build the definite integral of a function. On thinking over the definite integral decorator, I had almost decided that one necessary, but not sufficient, criterion for a good decorator is that it must not change a function's arg list. I was happy with this until just now. I've been fighting wxPython a lot recently, and suddenly a good working definition for a decorator came to me: * A decorator mediates between a function and its environment. In particular, I thought about something like: @mousexy def OnRightClick(self, x, y): ... For those non-wx'ers, all GUI events come wrapped in an "event", so all event-responding methods tend to look like: def OnRightClick(self, event): x = event.GetX() y = event.GetY() which looks like boilerplate to me. I'm wondering whether others think this is an interesting insight into what decorators are "for," or they think I'm working on "the moral equivalent of a macro." --Scott David Daniels Scott.Daniels(a)Acm.Org

3 2

McMillan Installer vs. Python 2.4
by mrmakent＠cox.net March 31, 2005

March 31, 2005

3 quick questions for the newsgroup: 1. Does anyone know why McMillan Installer 5b5 does not work with Python 2.4 under Linux (works with Python 2.3 just fine), and how to fix it? 2. Will anyone be picking up the maintenance and development ball for McMillan Installer? 3. Is there another, better-supported solution for distributing a Python executable under Linux/Unix/AIX? A single-file solution (ala Installer's '--onefile') is a requirement for me. 4. Installer supports the use of upx for those platforms that support it. AIX does not. Does anyone have any other solutions for compressing a Python executable which still leaves that executable executable? Believe it or not, the biggest impediment to my using Python at work is the enormous size of the executable. (Installing Python itself on the target platforms is not allowed). OK, that was 4.

3 4

Actor pattern in GUI
by ali.jan＠gmail.com March 31, 2005

March 31, 2005

Hi, I am trying to grok using actor patterns in a gui as explained here by Andrew Eland: http://www.andreweland.org/code/gui-actor.html The short article explains it using java with which i am not used to at all. But he does provide a python example using pygtk: http://www.andreweland.org/code/gui-actor.py Could anyone look into it and maybe explain it a bit as i don't really get at all what's happening? In addition I am getting these error when I try to run it: GLib-WARNING **: giowin32.c:1654: 3 is neither a file descriptor or a socket GLib-CRITICAL **: g_io_add_watch_full: assertion `channel != NULL' failed GLib-CRITICAL **: g_io_channel_unref: assertion `channel != NULL' failed Any ideas? Thanks...

1 1

Mac OS X Installer for Python 2.4.1 available
by Anthony Baxter March 31, 2005

March 31, 2005

Thanks to Bob Ippolito, there's now an installer for Python 2.4.1 available for Mac OS X 10.3 and later. Grab it from the Python 2.4.1 page - http://www.python.org/2.4.1/ Anthony

1 0

return the last item in a list
by David Bear March 31, 2005

March 31, 2005

I've googled for the above and get way too many hits.. I'm looking for an 'easy' way to have the last item in a list returned. I've thought about list[len(list)-1] but thought there would be a more gracefull way.

6 8

Pari Python
by Philp Smith March 31, 2005

March 31, 2005

Hi I'm interested to try Pari Python but haven't had much success in the past compiling such packages for Windows. Has anyone any advice or is there somewhere out there a precompiled version? Thanks Phil

1 0

Re: math - need divisors algorithm
by Philp Smith March 31, 2005

March 31, 2005

excellent - thanks in the meantime I managed to write something myself which a) works b) looks reasonably elegant and pythonic but c) is recursive d) requires a mainroutine and co-routine I have a feeling if I analysed it it would prove to be relatively inefficient as I'm sure it creates a lot of temporary objects (partial lists) So I will gratefully try your version - and for interest's sake post mine at some point (its on other computer) for comment. Phil "Tim Peters" <tim.peters(a)gmail.com> wrote in message news:mailman.1111.1112246542.1799.python-list@python.org... > [Philp Smith] >> Does anyone have suggested code for a compact, efficient, elegant, most >> of >> all pythonic routine to produce a list of all the proper divisors of an >> integer (given a list of prime factors/powers) > > If the canonical factorization of N is the product of p_i**e_i, the > number of divisors is the product of e_i+1. This can be > astronomically large, so it's important to minimize the number of > multiplications performed, and to write a generator to produce the > divisors (it may, e.g., be impossible to materialize a list of all of > them). > > This one's easy and efficient: > > def gen_divisors(prime_list): > elts = sorted(set(prime_list)) > numelts = len(elts) > > # Split on the number of copies of elts[i]. > def gen_inner(i): > if i >= numelts: > yield 1 > return > thiselt = elts[i] > thismax = prime_list.count(thiselt) > > # powers <- [thiselt**i for i in range(thismax+1)], > # but using only thismax multiplications. > powers = [1] > for j in xrange(thismax): > powers.append(powers[-1] * thiselt) > > for d in gen_inner(i+1): > for prime_power in powers: > yield prime_power * d > > for d in gen_inner(0): > yield d > > For example, do: > > for d in gen_divisors([2, 3, 2, 3, 2, 3, 5, 2, 3]): > print d > > You're on your own for the "proper" business; try not to use more than > one line to do it <wink>. > > I find recursion clearest here. An iterative version is easy enough > to do by incrementing from 0 thru product(e_i) (and just skipping the > endpoints if you really want "proper") in a mixed-radix system defined > by the e_i. Exercise for the reader. > > Harder: generate divisors in increasing order.

1 0

UnicodeEncodeError in string conversion
by Maurice LING March 31, 2005

March 31, 2005

Hi, I'm working on a script that searches a public database and retrives results using SOAP interface. However, it seems that the results may contains unicodes. When I try to pump the results into Firebird database using kinterbasdb module, some results will give me a UnicodeEncodeError. From what I've gathered, it looks like the conversion from the results from SOAP interface to string results in the error. Is there any way to get thru this? Thanks maurice

1 0

Weakrefs to classes that derive from str
by Ron Garret March 31, 2005

March 31, 2005

Why doesn't this work? >>> from weakref import ref >>> class C(str): pass ... >>> ref(C()) Traceback (most recent call last): File "<stdin>", line 1, in ? TypeError: cannot create weak reference to 'C' object >>> Note that this does work: >>> class D(int): pass ... >>> ref(D()) <weakref at 0x53e10; dead> >>> Likewise for floats, lists, etc. Everything but strs. rg

4 12