[sage-devel] Re: load/attach bugs

2009-04-27 Thread John Cremona
There is no need to put in a long description within the patch: just put in a short one and add detail on the trac ticket. And you can use whatever editor you like. For me hg pops up emacs, presumably because I have environment variable set to "emacs -nw" John 2009/4/27 Rado : > > Alright, her

[sage-devel] Re: load/attach bugs

2009-04-26 Thread William Stein
On Sun, Apr 26, 2009 at 5:59 PM, Rado wrote: > > Alright, here is my first open-source project patch: > > http://www.math.uiuc.edu/~rkirov2/sage/11803.patch Thanks. I've posted the patch here: http://trac.sagemath.org/sage_trac/ticket/5908 > > The description in the patch file is short becaus

[sage-devel] Re: load/attach bugs

2009-04-26 Thread Rado
Alright, here is my first open-source project patch: http://www.math.uiuc.edu/~rkirov2/sage/11803.patch The description in the patch file is short because vi is driving me insane. I was going to add a better description in the .patch file with gedit but wasn't sure if it is checksumed. Here is m

[sage-devel] Re: load/attach bugs

2009-04-25 Thread William Stein
On Sat, Apr 25, 2009 at 4:30 PM, Rado wrote: > > Sweet, I will give it a shot. > > One last thing I noticed is that attached_files() is not exposed to > the user. Again not sure if this is how its supposed to work or an > actual bug. But if one can attach without refering to a library it > would

[sage-devel] Re: load/attach bugs

2009-04-25 Thread Rado
Sweet, I will give it a shot. One last thing I noticed is that attached_files() is not exposed to the user. Again not sure if this is how its supposed to work or an actual bug. But if one can attach without refering to a library it would make sense to see what is currently attached again without

[sage-devel] Re: load/attach bugs

2009-04-24 Thread William Stein
On Fri, Apr 24, 2009 at 9:07 PM, Rado wrote: > > Alright I got how to make ``load "test space.py"`` work for the > notebook too. The problem is in: > > /home/rado/sage-3.4/devel/sage/sage/server/notebook/worksheet.py > > line 3558:  for filename in L.split(): > > the python split function splits

[sage-devel] Re: load/attach bugs

2009-04-24 Thread Rado
Alright I got how to make ``load "test space.py"`` work for the notebook too. The problem is in: /home/rado/sage-3.4/devel/sage/sage/server/notebook/worksheet.py line 3558: for filename in L.split(): the python split function splits "test space.py" to '"test' and 'space.py"'. I googled for a b

[sage-devel] Re: load/attach bugs

2009-04-24 Thread Nick Alexander
> I can fix 2) at $SAGE_ROOT/devel/sage/sage/misc/interpreter.py > but actually I need a fix for 1) since I keep my files with folders > with spaces. Since prompt works with files with spaces, it shouldn't > be too hard to make the notebook work with those too, right?!? Curiously the notebook han

[sage-devel] Re: load/attach bugs

2009-04-24 Thread Rado
I think I worded myself badly in the previous post. I see now that its a design decision to use this syntax USAGE: ``attach file1 file2 ...`` - space-separated list of .py, .spyx, and .sage files. instead of ``attach file1 arguments`` (this is the ipython syntax). You are right obviously

[sage-devel] Re: load/attach bugs

2009-04-24 Thread Nick Alexander
>> (a file to take arguments from command line) /home/rado/.sage/ >> testarg.py This is not what load/attach were intended for (and why do you expect it to work?), but it's not a terrible idea. Certainly attach should never support passing arguments, that's just perverse. Nick --~--~--

[sage-devel] Re: load/attach bugs

2009-04-24 Thread mabshoff
On Apr 24, 6:24 pm, Rado wrote: > Hello, Hi Rado, > I was trying to load some python files and found some weird behaviour > of the load/attach commands. I have three test files > > (a normal file) /home/rado/.sage/test.py > (a file with space in the name) /home/rado/.sage/test space.py > (a f