[Python-Dev] Import and unicode: part two

[Terry Reedy]

On 1/19/2011 7:34 AM, Victor Stinner wrote:
> Hi,
>
> I patched Python 3.2 to support modules with non-ASCII paths (*). It
> works well on all operating systems. But the task is not completly
> done:
>
> (a) Python 3 doesn't support non-ASCII module names (b) Python 3
> doesn't support unencodable characters in the module path
>
> I would like to know if we need to support that. Terry J. Reedy
> wrote (issue #10828): "I think bugs in core syntax should have high
> priority. I appreciate your work toward fixing it."

I am a little shocked at the so-far tepid response to (a), so let me
defend and explain my claim that it is a bug.

In the simplest case (from 6.11. The import statement and  2.3. 
Identifiers and keywords)

import_stmt ::= "import" module
module      ::= indentifier
identifier  ::= <appropriate Unicode start and continue chars>

There is nothing, nothing, about any restriction on identifiers.

The rest of 6.11 discusses the complex import algorithm but leaves out 
the simple semantics that cover 99% of cases (import a ???.py file in a 
directory on sys.path), and never mentions ".py".

So lets go to Tutorial 6. Modules which does explain the simple case: "A 
module is a file containing Python definitions and statements. The file 
name is the module name with the suffix .py appended." So, if xyz is a 
legal identifier and xyx.py exists on sys.path, it is reasonable from 
the docs to expect 'import xyz' to work. (Sys.path is memtioned in the 
reference.)

But we now have the following possibility:

Let xyz.py be

def double(x): return 2*x

if __name__=="__main__":
   if double(2) == 4: print("test passed")

We run the file, get "test passed", and write zyx.py:

import xyz
...

We run zyx and Python says "No module named xyz".

Bad, and quite puzzling to anyone who does not understand the subtle 
difference between running and importing a file.

-- 
Terry Jan Reedy