This doc improvement suggestion is inspired by #991196 (and subsequent duplicates) and the current discussion on py-dev in the thread
  'variable name resolution in exec is incorrect'
(which is not a correct claim). I believe there is consensus that the doc for exec needs improving.

My suggestion (which others may amend) is that the following paragraph (from the 3.x builtin functions exec entry)

"In all cases, if the optional parts are omitted, the code is executed in the current scope. If only globals is provided, it must be a dictionary, which will be used for both the global and the local variables. If globals  and locals are given, they are used for the global and local variables, respectively. If provided, locals can be any mapping object."

have these two sentences added:

"If only globals is provided or if onedict is provided as both globals and locals, the code is executed in a new top-level scope. If different objects are given as globals and locals, the code is executed as if it were in a class statement in a new top-level scope."

To be super-clear, consider adding one more sentence, something like

"The result for code with def statements or lambda expressions may be different than it would be if the implied context were a function rather than a class."

I'm going to update the documentation to include the scope information. Should be done within a day.

Patches for 2.x and 3.x documentation related to 'exec'.

2.x
  Provides more information about scopes when using additional expressions
  for 'exec'. 2.x documentation for 'exec' is found in
  reference/simple_stmts.
3.x
  Provides more information about scopes when using additional expressions
  for 'exec'.  3.x documentation for 'exec' is found in library/functions.

In 3.x exec is a function, so the reference to 'in' should be removed/updated.  On 2.x it might be better to just use ``in`` instead of :keyword:`in`, because the latter probably links to the 'in' operator that checks for containment.

I've updated the 3.x patch, should be correct now.

I also updated the 2.x patch to use ``in``. :keyword:`in` was used in an earlier part of the paragraph I modified, so I have changed both for consistency.

This is the paragraph before modification:
"
  In all cases, if the optional parts are omitted, the code is executed in the
  current scope.  If only the first expression after :keyword:`in` is specified,
  it should be a dictionary, which will be used for both the global and the local
  variables.  If two expressions are given, they are used for the global and local
  variables, respectively. If provided, *locals* can be any mapping object.
"

Should this be closed as a duplicate of issue #13557?

I suggested more change here than I committed in #13557. I would like to think about the extra a bit more. The issue about def in exec just came up again on python-list, but I do not really like the third sentence I suggested.

Okay;  sorry about that.  Reopening.

I decided that the alternate addition in #13557 is enough.

resolution: out of date
stage: patch review -> resolved

type: enhancement
versions: + Python 3.4, - Python 3.1

messages: + msg135820

stage: patch review