| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Comments:
<0> ok, then i missed something, unless someone defined partial themselves ;)
<1> but type(some_func) <type curried, blah> looks like an interpreter stuff
<2> Yango: Jerub's example was from MochiKit, which is a javascript library
<3> (based on pythonic principals)
<2> I don't know whether Python has a partial or not, offhand.
<3> TML: it will.
<3> anyway, the implemntation of partial looks like:
<3> http://rafb.net/paste/results/A0gmLK18.html
<4> beautifulsoup++
<0> masala, hehe
<0> a perhaps clearer version (doesn't do kwargs and only curries one arg) is something like:
<5> yeah, beautifulsoup is really cool.
<0> >>> def partial(f, firstarg):
<0> ... def f2(*args):
<0> ... return f(firstarg, *args)
<0> ... return f2
<1> i love that straightness of python. I still can't get used to the idea of messing with args and functions like that
<2> Jerub: Does that actually happen often enough that it warrants language-level support?
<0> did you know you can put functions on functions?
<0> Jerub: i think it will be in a module
<3> TML: it's only getting put into the builtins.
<1> mcmillen, yes
<0> erm, TML
<3> TML: and it was to avoid people getting pissed off about lambda going away, which it isn't now.
<5> wait, it's not?
<2> I see
<0> yango: i actually used that feature the other day, for a @memoize decorator
<0> gzl: nope
<5> ....
<5> cool.
<3> gzl: no, guido got pissed off with people wasting their time convincing him, and wasting their time inventing a better syntax.
<5> I see.
<1> that's another topic widely misstudied (and thus misunderstood) by me :) Python's decorators aren't exactly the GOF pattern implementation, right?
<5> Jerub: wait, so he's just leaving it as is, or is using some new syntax?
<3> gzl: it's staying exactly as it is.
<5> ok
<3> Yango: nope
<0> yango: not really.
<4> but it's still going away, right?
<3> Yango: the GOF implementatino was only designed the way it was to get around language limitations.
<3> Leeds: no, it's not.
<0> Leeds: no
<6> :)
<1> Jerub, the objective though is the same?
<3> Yango: dunno, I've ignored the GOF stuff since I came to python
<2> Jerub: I'm still having trouble divorcing myself from my old internals@php.net attitude, where adding stuff to the core is what got us in the trouble we're in. :)
<3> Yango: most of the patterns in their catalog in the book are designed to combat language limitations.
<7> I never got the point of singletons
<7> if you only want an object once, why not just instantiate it once?
<3> steg: again, language limitation.
<3> steg: in java, you can't do:
<3> cl*** Foo: p***
<3> Foo = Foo()
<3> and be done with it.l
<1> The ideas are valuable though, regardless of language limitation
<3> also, very few languages have this idea that the programmer is a consenting adult.
<3> Yango: oh, sure.
<7> Jerub: Ahh.
<1> so, decorators in python append generic functionality to a function
<1> ?
<2> Jerub: I disagree. I always felt the point of the patterns book was to codify and name common problems faced in software. Solution implementation was a secondary goal, and only broadly discussed.
<0> decorators are syntactic sugar for a function that operates on a function
<3> TML: I ignore the patterns catalog.
<3> TML: but I love the first section.
<8> the only decorater ive seen and kind of groked was a @sychronised one
<3> TML: but they don't publish a version of the book without the catalog.
<9> how would i match an underscore with a regex?
<3> _
<1> but operates on a function result, args, anything?
<8> with _
<3> parks: seen turbogears ?
<3> parks: it uses decorators extensively.
<0> yango: @someDecorator def foo(): p*** is *equivalent* to simply def foo(): p*** foo = someDecorator(foo)
<9> Jerub: it doesn't seem to work with a string like ' "_ '
<8> Jerub no
<3> mariodemon_11: are you using match or search?
<10> The term 'decorator' in Python refers to a method wrapping construct, not to the term 'Decorator' in the gang of four.
<3> parks: watch the 20 minute introductory screenscast.
<3> parks: it's 30mb, available from turbogears.org
<0> Yango: parks: see http://pastebin.com/577755 for a definition of a simple decorator
<9> Jerub: o damn, i forgot, thanks
<0> that "memoizes" a function... i.e. if the function has been called with the same args before, it returns the cached value
<1> I love how geeks love to invent terms... :) cache_result seems a lot clearer to me
<0> you could say something like def fib(n): if n <= 2: return 1 else: return fib(n-1) + fib(n-2), then say fib = memoized(fib). or you can say @memoized before the def of fib, which is clearer
<0> memoized isn't my word, i blame Guy Steele :)
<8> why would fib get called with the arguments
<1> I wasn't blaming you, mcmillen
<8> the same arguments
<8> that should read
<1> fib(n-1) was called before as fib(n), parks
<1> or seomthing like that :)
<1> I'm not thinking straight
<0> parks: run my def of fib() without memoziing it and see how long it takes
<11> is it possible to have an object emulate a boolean? I have a singleton NotSet which I would like to behave as None, so that NotSet or b == b
<1> memoiizing, ell oh ell
<12> http://rafb.net/paste/results/GytlWX75.nln.html'
<0> to compute fib(100), it would have to compute fib(99) and fib(98), but fib(99) also wlill recompute fib(98), and the computational time blows up exponentially
<12> anyone have a clue why that doesn't connect?
<0> with memoization, you compute fib(98) just once, then return it immediately everywhere else that it's called
<0> admittedly, fib is a dumb example because it has a trivial iterative method of computing it, but for some functions recursion is nice :)
<12> no one?
<0> kjetilho: define a __nonzero__ method
<11> mcmillen: ah. I'm stuck with Python 2.3, but it seems I can use __len__ instead for this purpose
<0> kjetilho: i think nonzero has existed since 2.2, but let me check
<11> you're right
<11> sorry, thanks a lot!
<0> np! enjoy ;)
<8> this channel is the best for procastination
<1> parks++
<8> bulio put some print statements after each call to see whats happening
<1> and you even learn some things!
<8> kjetilho whats wrong with True, False and None ?
<12> parks: ok
<0> yango: you mean parks += 1? :)
<8> zanella thinking about doing work, but doing everything to avoid doing said work
<11> parks: I use it for keyword arguments where True, False and None are valid values
<1> i mean operator.add(parks,1)
<8> is there any value to being incremented?
<13> parks: thanks
<11> parks: I guess it wouldn't be an issue if there weren't an inheritance tree
<0> yes, if you reach sys.maxint you progress to Level 2 of Python Mastery.
<14> mcmillen: don't you just roll over to the negatives? O.o
<0> roderyk: possibly, but that's also pretty cool :)
<3> then, once you manage to fill the entire amount of memory a python process can hold with your number, you trancend, and you get ops in #python
<0> i really only care about abs(mcmillen)
<0> so if i ever go <0, i'll become a trolling flamer
<0> you guys should be lucky i said somethin ginsightful my first time :D
<12> how do I run python from windows command line?
<14> the issue I'm worried about is if we start comparing each other relatively
<12> I can't get my script running
<12> its been awhile since I've used windows
<14> so parks/mcmillen which will lead to floats...
<3> bulio|: C:\python24\python.exe
<3> or whatever
<3> just like you'd run any other program.
<14> I want to be a _whole_ person ><
<12> Jerub: but to run a script
<12> not python itself
<3> bulio|: C:\python24\python.exe myscript.py
<13> bulio| You don'treally like to read docs do you?
<15> hello. I just wrote a small wrapper that I think is useful, in python+pygtk+glade. I only need the .py file and .glade to use this, but how should I go about packaging my app to make it installable?
<1> engla, tar cvzf installer.tgz file.py file.glade :-) (you mean under windows?)
<16> This is more difficult than I thought.
<12> py2exe?
<12> I can run python
<12> but not the script
<15> Yango: linux
<15> but okay.. if that's so. But it would be nice to be able to make a .deb
Return to
#python
or
Go to some related
logs:
#ubuntu
#perl
#redhat
+trs-80 +oregon trail
#web
sanderd d1
#ubuntu
hibernate nested sets
libvlcplugin undefined symbol xtwindowtowidget
#perl
|
|