Python Project Architecture

import from __future__ allows you to access new features which are not currently compatable with the current interpreter

import __future__ module allows you to see when features were added and became default with __future__.<feature>

In python, .py files are modules of definitions and statements

in a module, __name__ is the name of the module
statements in a module are executed during the import statement, used for module initialization
each module has a private symbol table (sheltered from global variable scope)
from <module> import * will import names from a module into the importing module's symbol table
- names beginning with _ are not imported
if __name__ == "__main__" allows you to make a module both a usable script and a module to be imported
search order for imports: built-in module, .py file in sys.path
dir() is used to find out which names a module defines

Packages provide namespacing for modules, like A.B for package A and submodule B

__init__.py initializes a package (in its current directory, based on the directory name)
packages may be nested, to namespace imports, but every item except the last must be a package in statements like import item.subitem.subsubitem
in __init__.py, a list named __all__ is the list of submodules to be imported when from <package> import * is called
within a package, relative imports can be made with from .. import <submodule>

tensorflow directory defines the package
- __init__.py imports 3.x absolute_import, division and print_function from __future__
- from tensorflow.python import *, which imports the list defined by __all__ in the tensorflow.python package
  - python/__init__.py imports from tensorflow.core, tensorflow.framework, and sets default environment variables, and global default config