Why I don't need to declare encoding when using python interpreter? -

# -*- coding: utf-8 -*- 

i understand line of code necessary when non-ascii characters involved in python script file.

when learning python, told 2 ways of running python code (line line in interpreter vs run script file) yield same result. , do, in cases. when non-ascii characters involved in scripts, turns out have declare encoding first.

moreover, have tried exec() function, trying execute string containing python codes.

>>> exec ("b='你'") 

it works.

but if save "b = '你'" script , run it, syntax error.

i curious why don't need declare encoding when running python codes line line in interpreter.

is there difference in executing procedures of these 2 way?

thank you.

i suppose interactive session of python use system encoding (see python unicode strings , python interactive interpreter).

when read source file, need know how interpret data it's parsing it's logical script not necessary written in same encoding terminal executing it; , more when script ran without environment specifying encoding.

also, interesting read joel spolsky the absolute minimum every software developer absolutely, positively must know unicode , character sets might explain why python choose require developers explicit encoding (which have preferred set default utf8, way coherent zen of python).