[Tutor] Confused about globals

Dustin Mitchell dustin at ywlcs.org
Fri Jun 9 16:19:42 CEST 2006


As a note, the "import" should be

   import shared_funcs

In Python, most globals aren't really global -- they're local to the 
module.  If you split your modules by functionality, then variables 
should naturally relate to a specific module, and be located there.  
So, for example, you could move those global variables to 
shared_funcs.py, renamed to wells.py:

=== wells.py ==

DE={}
DN={}

def get_xy_data():
	in_file=open("well_xy.txt","r")
	for line in in_file
		L=line.split()
		well=L[0]
		x=L[1]
		y=L[2]

		DE[well]=x
		DN[well]=y
	in_file.close()

def lookup_well(well):
   return (DE.get(well, None), DN.get(well, None))

=== main.py ===

import wells

def func_that_needs_wells():
   ...
   for e n wells.DE.keys():
     ...
   ...

etc.

On Jun 9, 2006, at 4:42 AM, Etrade Griffiths wrote:

> Hi
>
> I have a series of python programs that plot stuff using PYX.  Common 
> to
> these is the need to read in a list of (well) locations in (X,Y) 
> coords so
> I put that code in a separate module called shared_funcs.py.  The 
> coords
> are stored in dictionaries which I want to use later in the "main" 
> program
> - every time I find a well with data, I can get the well's (x,y) 
> coords by
> looking up the well name in the dictionary.  The code snippet looks 
> like this:
>
> === main.py ====
>
> # define coord dictionaries for global use
>
> DE={}
> DN={}
>
> import shared_funcs()
>
> shared_funcs.get_xy_data()                 # Get coords from file
>
>   ... do plotting stuff
>
>
> === shared_funcs.py ===
>
> def get_xy_data():
> 	in_file=open("well_xy.txt","r")
> 	for line in in_file
> 		L=line.split()
> 		well=L[0]
> 		x=L[1]
> 		y=L[2]
>
> 		DE[well]=x
> 		DN[well]=y
> 	in_file.close()
>
> The problem is that DE and DN appear not to be accessible to 
> get_xy_data -
> presumably this is because shared_funcs.py and main.py don't share the 
> same
> scope.  So, is there any way to let get_xy_data change DE and DN?  I 
> guess
> the obvious way is to pass them to get_xy_data as arguments - but is 
> there
> a more "pythonic" method?  Thanks in advance!
>
>
> _______________________________________________
> Tutor maillist  -  Tutor at python.org
> http://mail.python.org/mailman/listinfo/tutor



More information about the Tutor mailing list