| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Comments:
<0> :(
<1> pythonwin = weird behavior
<0> sure does :(
<2> * + firefox = weird behaviour
<1> i use emacs myself, tho i also have SPE sitting around not doing anything
<2> firefox = weird behaviour
<2> :P
<3> Hello all! What would be the "best" way to find the dotted path of an objects cl***?
<3> object.__cl***__.__module__ + extracting the cl*** name for object.__cl***__, or some other way?
<3> I googled but couldn't find anything.
<2> the dotted path?
<3> Yeah, like datetime.datetime
<3> Or mymodule.something.foo.MyCl***.
<2> look in sys.modules?
<3> Ehm, but thats a list of modules, isn't it?
<4> is there a way to use sorted() in python2.3 -- i.e. some code i can copy? right now i just defined a function sorted(l): l2=copy.copy(l); l2.sort(); return l2
<2> list of loaded modules
<5> regebro: twisted has a function to do this you could look at: twisted.python.reflect.qualify()
<5> regebro: i don't think it works particularly well
<1> modules don't have a path hardwired into them like java cl***es do. the path only resolves the module, but the module only has an unqualified name
<1> s/cl***es/packages/
<6> mcmillen_: well, that's basically it
<3> joedj: OK, I'll check it out. Do you know if there is any specific problems with it?
<6> except maybe you just would do l2 = list(l)
<6> to support any iterable
<6> and add those keyword arguments
<6> mcmillen_: it won't be as fast as sorted() - which I think is optimized for this kind of "insertion sort" technique
<3> sproingie: Well, in this case I don't care about modules, just objects.
<5> regebro: i don't recall
<4> all right. mostly i was wondering if the actual function is written in python somewhere, so that i could just copy the source. but if it's written in C, i'll just stick with what i have
<3> ok, thx.
<6> mcmillen_: I'm afraid it's written in C, but you might find an earlier Python in the PEPs
<1> regebro: then you're even more out of luck. objects have no idea what module they belong to
<1> regebro: technically they belong to every module that imports them
<3> object.__cl***__.__module__ seems to work for me.
<4> stain: thanks for the help!
<3> sproingie: Ah, OK, now I understand.
<7> sproingie: Well, you could probably localize code objects, and access the "globals", but that's really an ugly hack.
<3> That's not a problem per se. As long as I can import it (which I can) I'm fine.
<3> I think. :)
<1> ok, i'm wrong again, big surprise ... __module__ indeed does it. not going to give you qualified names tho.
<8> if you already have a reference to the cl***, why would you need to reimport its module?
<3> I don't have a reference to the cl***. I have an object.
<1> object.__cl***__
<8> regebro you have a name of an object, which is a reference to that object
<8> the unnamed object is just floating around in memory
<3> No, I have an OBJECT. Not teh name. The OBJECT.
<8> I don't believe you
<3> joedj: twisted just does cl***.__module__ + '.' __cl***__.__name__
<3> Zalamander: Beleive it.
<8> regebro no
<8> regebro I think you are misunderstanding something about names/objects in Python
<3> OK, so explain.
<8> regebro "objects" exist in memory. "names" are references to objects. You can't "have" an object in a python program unless you have at least one "name", i.e., reference, for it
<1> last i looked names were strings
<8> no they aren't
<1> what an amusing bit of epistemology
<8> but the real point is not what names are ... the real point here is "why import a cl*** that you already have a reference to?"
<3> Zalamander: No, "names" are not referencce, and everythins references. I have a reference, i.e. an object.
<3> What I need, is the import path of the object.
<3> As a string.
<8> regebro yes, I understand what you need. I only asked why.
<1> you can get the unqualified name of the module, but i think the qualified path gets thrown away
<3> No, Zalamander, you told me I had misunderstood it. ;) That's not asking why. ;)
<9> Zalamander: You may have a reference but not know where it is
<8> regebro <8> if you already have a reference to the cl***, why would you need to reimport its module?
<2> m0no: use opera :P
<8> that was the first thing I said to you regebro
<1> at least it does in the module you imported. you could probably grovel sys.modules to match them up
<3> I'm not qure I understand the difference between qualified and unqualified in this context.
<8> Brend agreed. That's why I asked why one would need to know?
<1> regebro: qualified: foo.bar.modulename. unqualified: modulename
<3> The reason why is an export/import thingie.
<8> regebro but I do maintain that you don't "have" the object. You "have" a reference to the object (Unless you can somehow visually interpret electronic impulses in capacitors)
<3> Snce everything in python is references, that is a compltely useless distinction.
<8> regebro ok, I don't understand that reason
<8> regebro no, it isn't. Not in this case
<3> It's fine, you don't have to. My question has been answered, and yes, it is.
<8> it's exactly the important distinction. You can export something from the CURRENT namespace, without knowing where the cl*** is defined
<8> because you have a LOCAL reference to it
<3> I will not have a local reference to it.
<1> Zalamander: no one is impressed by your ability to distinguish references from values
<3> It's not a problem.
<8> then you won't be able to look it up
<1> Zalamander: in python, "object" is synonymous with references. you're correcting someone for confusion that doesn't exist
<8> sproingie hey, **** off :-)
<3> Yes, I will be able to look it up. In fact, I just did.
<1> same thing happens in java too, tho there some people actually do expect p***-by-value to happen for some ungodly reason
<1> anyway, object.__cl***__.__module__ gives you the name of the module (very much a string), and sys.modules has the qualified names and the module objects ... beats hell out of me how you can get a reference to the module of an arbitrary cl***
<10> hi
<3> 'arbitrary'?
<1> arbitrary, given, anonymous, whatever
<1> long as you don't have any overlapping module names, it's just a matter of matching up the module name with the keys in sys.modules
<1> if they do overlap ... good luck
<3> Well, well see, but I don't think this is going to be a problem, really.
<1> it's a one-liner code exercise i'm too lazy to figure out
<11> Is there an easy way to create an html-file with syntax highlighted python code?
<1> sp3tt: enscript
<1> sp3tt: http://www.codento.com/people/mtr/genscript/
<11> Found a plugin for jEDit.
<1> jedit++
<1> could use a new maintainer, considering its author has sworn off java
<11> Nice, the plugin generates valid html, except for a doctype.
<1> how hard is it to generate valid html for colors and fonts?
<12> sproingie, fairly easy
<13> anyone here works with pygtk?
<14> yep
<13> Juhaz, do you know how can i have 2 windows but dont let the focus go to the back window till you click OK on the first?
<15> sorted as raw strings, not alphanumerics <-- what is raw string ?
<14> sapo, um, I'm sure there are ways to accomplish that entirely by hand, by if you need a modal dialog, why don't you just use one?
<13> hum.. modal
<13> i m new to it.. i m using glade.. its working but i m kinda lost with some stuff
<16> sapo, new to Python?
<13> kinda
<13> i m playing with it a lot latelly :)
<13> now i m trying to convince my boss that python is worth for desktop apps.. i m making some server side scripts in python and they are working fine, but i m starting with gui stuff
<17> what are they currently considering sapo:
<13> just for testing i m trying to make a app that asks for your user and p***word, this login window is working
<14> http://pygtk.org/pygtk2reference/cl***-gtkdialog.html#constructor-gtkdialog http://pygtk.org/pygtk2reference/cl***-gtkwindow.html#method-gtkwindow--set-modal
<13> but i want to lock the focus on it till you login
<18> solsTiCe_: A raw string is a string literal prefixed with r which makes the parser not use \ as a escape char. That term doesn't apply to your sentence though.
<13> Juhaz, thanx i ll take a look
<15> Erwin: ok
<19> sapo: what os and platform requirements for your gui stuff offhand?
<14> sapo, I don't have glade at hand right now, but pretty sure that the gtk.DIALOG_MODAL flag can also be set trough it
<17> sapo: What are they planning to use for the gui ?
<13> Juhaz, thanx man.. this worked like a charm: Tree.get_widget("login").set_modal(True)
<13> JoaoJoao, hum.. we actually make commercial applications for small companies using delphi.. but i hate delphi.. and we are starting to make a new project and i m in that project.. i dont wanna make that stuff in delphi
<17> sapo: You may have a hard time convincing them to dump the Delphi's rad
<13> JoaoJoao, i think i ll, but i have some credits with my boss.. everything i said till now worked, so he will consider if i show him something that already works.. thats why i m making an example right now :)
<20> sapo, I've heard people combining Python and Delphi to good effect; doing UI dev in Delphi and the "guts" in Python
<17> yes there is a Python component for Delphi
<13> benji, hum... never heard about this combination O_o
<13> what i really hate about delphi is that object pascal stuff
<18> Or you can skip Delphi altogether and use e.g. dabo, which is a 3-tier Python-based toolkit
<13> delphi itself is a good rad, but that language give me nighmares
<13> dabo?
<17> Delphi is dying
<17> Delphi 2005 ****s
<13> JoaoJoao, yep.. borland is even selling it.. lol
<3> Pah, just because it's an ugly shim with some parts undocumented= That's not reason to hate it. :)
<17> I'd rather use VisualStudio, it's nice
<21> Hi :)
<17> regebro: And very slow, and buggy, did I mention it's slow? :)
<13> Ghawk, yo
<3> I like how object pascal tries to pretent that you don't need pointers. It's sooo...pathetic. :-P
<17> what saves delphi is the wealth of libraries and components for it
Return to
#python
or
Go to some related
logs:
#openzaurus
uchromep linux
#linux
Gtk-WARNING **: cannot open display: in linux CentOS 5.0
#centos
#mysql
up2date cups bus error
richzo
keymap evbug
xorg gcc: cannot specify -o with -c or -S and multiple compilations
|
|