Actually, that clears up a lot - thanks.
What you are saying is that the def statements are executable, and in fact, their execution is causing modification to the execution environment in such a way as to cause the function to be defined for interpreted code from that point forward. The "def" is actually an "inject the following function into the symbol table". This is to say that a function call is recognized as such by virtue of its syntax e.g., foo(bar,baz), basically turns into call_function( "foo", (blah,baz) ), and that the name itself isn't dealt with until the function call actually occurs. There would be nothing special about the name "foo" when the interpreter first encounters it. With that in mind, I suppose that I could write some other function, which returns some compiled code, and then add that to my symbol table as "foo" and then hit some code that calls it. Hmm. > > [EMAIL PROTECTED] wrote: > > > As a bit of a [python] newbie myself, the proper procedure for forward > > declaring things has so far evaded me as well. > > > > Is there a "right" way to do this in python? : > > ... > > > > def f1( i ): > > if( i < 20 ): return f2(i) > > else: return 0 > > > > def f2( i ): > > if( i > 20 ): return f1(i) > > else: return 0 > > This code will work just fine. > > The trick to remember, is that the def statement is an executable statement. > The def statement for any function must be run by the interpreter, before the > function can be called. But the interpreter doesn't look up names until > runtime, so you can use the *name* of the function before you define it -- > it's all dependent on the order that the interpreter processes things. > > In the code above, when this file is run/imported, the interpreter will read > through this, see the def statements, and create two function objects. > Later, when you actually call f1(5), the interpreter steps throught the body > of f1(), sees that the argument passes the if-test, then sees the name f2. > At this point, the interpreter does a name-lookup to find what f2 is bound > to. If the interpreter hadn't already seen 'def f2(i): ...', then we'd get > the NameError now. In this case, it sees "oh, yes, I have this function > that's named f2..." and proceeds to call the function. > > Hope this makes things clearer... :) > > Jeff Shannon > Technician/Programmer > Credit International > > > _______________________________________________ > ActivePython mailing list > [EMAIL PROTECTED] > To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs -- _______________________________________________ ActivePython mailing list [EMAIL PROTECTED] To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
