[Tutor] [Long]General Questions

[Jeff Shannon]

gnux at freesurf.fr wrote:

>I've got a question concerning this prog: why have u written
>"radius=input()"rather than "radius=raw_input()"
>I thought we use the second when we ask something to the user
>Thx if u can explain me the difference between the two and why u
>use the second
>

Both input() and raw_input() prompt the user to type something, and 
return that to the program.  However, they handle what's typed a bit 
differently.

raw_input() will always return a string, exactly as the user typed it. 
 input(), on the other hand, evaluates that string as if it were a 
Python expression -- input() is equivalent to eval(raw_input()).  This 
makes input() more convenient for entering numbers, because you can just 
type the number and get an integer (or float, or whatever), just as if 
you'd typed a numeric literal in the interpreter.  By comparison, with 
raw_input(), you'd then have to use int() or float() to convert the 
string into a number.  On the other hand, input() is less convenient for 
entering strings, because (just like a string literal in the 
interpreter) you need to use quotes around it.  And if the quotes are 
forgotten, then whatever is typed is interpreted as a Python identifier, 
which might look up an existing variable or might throw a NameError. 
 And since input() effectively executes an arbitrary expression, it 
leaves you *wide* open to malicious (or accidental) attacks on your 
computer system.

In short, input() should only be used when you know exactly what's going 
to happen.  Which, really, should only be when you're working in the 
interactive interpreter.  It's too dangerous for routine production use.

Jeff Shannon
Technician/Programmer
Credit International