29.6. importlib – An implementation of import

3.1 新版功能.

29.6.1. Introduction

The purpose of the importlib package is two-fold. One is to provide an implementation of the import statement (and thus, by extension, the __import__() function) in Python source code. This provides an implementation of import which is portable to any Python interpreter. This also provides a reference implementation which is easier to comprehend than one in a programming language other than Python.

Two, the components to implement import can be exposed in this package, making it easier for users to create their own custom objects (known generically as an importer) to participate in the import process. Details on providing custom importers can be found in PEP 302.

参见

The import statement

The language reference for the import statement.

Packages specification

Original specification of packages. Some semantics have changed since the writing of this document (e.g. redirecting based on None in sys.modules).

The __import__() function

The built-in function for which the import statement is syntactic sugar.

PEP 235

Import on Case-Insensitive Platforms

PEP 263

Defining Python Source Code Encodings

PEP 302

New Import Hooks.

PEP 328

Imports: Multi-Line and Absolute/Relative

PEP 366

Main module explicit relative imports

PEP 3120

Using UTF-8 as the Default Source Encoding

29.6.2. Functions

importlib.__import__(name, globals={}, locals={}, fromlist=list(), level=0)

An implementation of the built-in __import__() function. See the built-in function’s documentation for usage instructions.

importlib.import_module(name, package=None)

Import a module. The name argument specifies what module to import in absolute or relative terms (e.g. either pkg.mod or ..mod). If the name is specified in relative terms, then the package argument must be set to the name of the package which is to act as the anchor for resolving the package name (e.g. import_module('..mod', 'pkg.subpkg') will import pkg.mod).

The import_module() function acts as a simplifying wrapper around importlib.__import__(). This means all semantics of the function are derived from importlib.__import__(), including requiring the package from which an import is occurring to have been previously imported (i.e., package must already be imported). The most important difference is that import_module() returns the most nested package or module that was imported (e.g. pkg.mod), while __import__() returns the top-level package or module (e.g. pkg).

29.6.3. importlib.abc – Abstract base classes related to import

The importlib.abc module contains all of the core abstract base classes used by import. Some subclasses of the core abstract base classes are also provided to help in implementing the core ABCs.

class importlib.abc.Finder

An abstract base class representing a finder. See PEP 302 for the exact definition for a finder.

find_module(fullname, path=None)

An abstract method for finding a loader for the specified module. If the finder is found on sys.meta_path and the module to be searched for is a subpackage or module then path will be the value of __path__ from the parent package. If a loader cannot be found, None is returned.

class importlib.abc.Loader

An abstract base class for a loader. See PEP 302 for the exact definition for a loader.

load_module(fullname)

An abstract method for loading a module. If the module cannot be loaded, ImportError is raised, otherwise the loaded module is returned.

If the requested module already exists in sys.modules, that module should be used and reloaded. Otherwise the loader should create a new module and insert it into sys.modules before any loading begins, to prevent recursion from the import. If the loader inserted a module and the load fails, it must be removed by the loader from sys.modules; modules already in sys.modules before the loader began execution should be left alone. The importlib.util.module_for_loader() decorator handles all of these details.

The loader should set several attributes on the module. (Note that some of these attributes can change when a module is reloaded.)

• __name__

  The name of the module.

• __file__

  The path to where the module data is stored (not set for built-in modules).

• __path__

  A list of strings specifying the search path within a package. This attribute is not set on modules.

• __package__

  The parent package for the module/package. If the module is top-level then it has a value of the empty string. The importlib.util.set_package() decorator can handle the details for __package__.

• __loader__

  The loader used to load the module. (This is not set by the built-in import machinery, but it should be set whenever a loader is used.)

class importlib.abc.ResourceLoader

An abstract base class for a loader which implements the optional PEP 302 protocol for loading arbitrary resources from the storage back-end.

get_data(path)

An abstract method to return the bytes for the data located at path. Loaders that have a file-like storage back-end that allows storing arbitrary data can implement this abstract method to give direct access to the data stored. IOError is to be raised if the path cannot be found. The path is expected to be constructed using a module’s __file__ attribute or an item from a package’s __path__.

class importlib.abc.InspectLoader

An abstract base class for a loader which implements the optional PEP 302 protocol for loaders that inspect modules.

get_code(fullname)

An abstract method to return the code object for a module. None is returned if the module does not have a code object (e.g. built-in module). ImportError is raised if loader cannot find the requested module.

get_source(fullname)

An abstract method to return the source of a module. It is returned as a text string with universal newlines. Returns None if no source is available (e.g. a built-in module). Raises ImportError if the loader cannot find the module specified.

is_package(fullname)

An abstract method to return a true value if the module is a package, a false value otherwise. ImportError is raised if the loader cannot find the module.

class importlib.abc.PyLoader

An abstract base class inheriting from importlib.abc.InspectLoader and importlib.abc.ResourceLoader designed to ease the loading of Python source modules (bytecode is not handled; see importlib.abc.PyPycLoader for a source/bytecode ABC). A subclass implementing this ABC will only need to worry about exposing how the source code is stored; all other details for loading Python source code will be handled by the concrete implementations of key methods.

source_path(fullname)

An abstract method that returns the path to the source code for a module. Should return None if there is no source code. ImportError if the module cannot be found.

load_module(fullname)

A concrete implementation of importlib.abc.Loader.load_module() that loads Python source code. All needed information comes from the abstract methods required by this ABC. The only pertinent assumption made by this method is that when loading a package __path__ is set to [os.path.dirname(__file__)].

get_code(fullname)

A concrete implementation of importlib.abc.InspectLoader.get_code() that creates code objects from Python source code, by requesting the source code (using source_path() and get_data()), converting it to standard newlines, and compiling it with the built-in compile() function.

get_source(fullname)

A concrete implementation of importlib.abc.InspectLoader.get_source(). Uses importlib.abc.ResourceLoader.get_data() and source_path() to get the source code. It tries to guess the source encoding using tokenize.detect_encoding().

class importlib.abc.PyPycLoader

An abstract base class inheriting from importlib.abc.PyLoader. This ABC is meant to help in creating loaders that support both Python source and bytecode.

source_mtime(fullname)

An abstract method which returns the modification time for the source code of the specified module. The modification time should be an integer. If there is no source code, return None. If the module cannot be found then ImportError is raised.

bytecode_path(fullname)

An abstract method which returns the path to the bytecode for the specified module, if it exists. It returns None if no bytecode exists (yet). Raises ImportError if the module is not found.

write_bytecode(fullname, bytecode)

An abstract method which has the loader write bytecode for future use. If the bytecode is written, return True. Return False if the bytecode could not be written. This method should not be called if sys.dont_write_bytecode is true. The bytecode argument should be a bytes string or bytes array.

29.6.4. importlib.machinery – Importers and path hooks

This module contains the various objects that help import find and load modules.

class importlib.machinery.BuiltinImporter

An importer for built-in modules. All known built-in modules are listed in sys.builtin_module_names. This class implements the importlib.abc.Finder and importlib.abc.InspectLoader ABCs.

Only class methods are defined by this class to alleviate the need for instantiation.

class importlib.machinery.FrozenImporter

An importer for frozen modules. This class implements the importlib.abc.Finder and importlib.abc.InspectLoader ABCs.

Only class methods are defined by this class to alleviate the need for instantiation.

class importlib.machinery.PathFinder

Finder for sys.path. This class implements the importlib.abc.Finder ABC.

This class does not perfectly mirror the semantics of import in terms of sys.path. No implicit path hooks are assumed for simplification of the class and its semantics.

Only class method are defined by this class to alleviate the need for instantiation.

classmethod find_module(fullname, path=None)

Class method that attempts to find a loader for the module specified by fullname on sys.path or, if defined, on path. For each path entry that is searched, sys.path_importer_cache is checked. If an non-false object is found then it is used as the finder to look for the module being searched for. If no entry is found in sys.path_importer_cache, then sys.path_hooks is searched for a finder for the path entry and, if found, is stored in sys.path_importer_cache along with being queried about the module. If no finder is ever found then None is returned.

29.6.5. importlib.util – Utility code for importers

This module contains the various objects that help in the construction of an importer.

importlib.util.module_for_loader(method)

A decorator for a loader method, to handle selecting the proper module object to load with. The decorated method is expected to have a call signature taking two positional arguments (e.g. load_module(self, module)) for which the second argument will be the module object to be used by the loader. Note that the decorator will not work on static methods because of the assumption of two arguments.

The decorated method will take in the name of the module to be loaded as expected for a loader. If the module is not found in sys.modules then a new one is constructed with its __name__ attribute set. Otherwise the module found in sys.modules will be passed into the method. If an exception is raised by the decorated method and a module was added to sys.modules it will be removed to prevent a partially initialized module from being in left in sys.modules. If the module was already in sys.modules then it is left alone.

Use of this decorator handles all the details of which module object a loader should initialize as specified by PEP 302.

importlib.util.set_loader(fxn)

A decorator for a loader method, to set the __loader__ attribute on loaded modules. If the attribute is already set the decorator does nothing. It is assumed that the first positional argument to the wrapped method is what __loader__ should be set to.

importlib.util.set_package(fxn)

A decorator for a loader to set the __package__ attribute on the module returned by the loader. If __package__ is set and has a value other than None it will not be changed. Note that the module returned by the loader is what has the attribute set on and not the module found in sys.modules.

Reliance on this decorator is discouraged when it is possible to set __package__ before the execution of the code is possible. By setting it before the code for the module is executed it allows the attribute to be used at the global level of the module during initialization.

29.6.6. Example

Below is an example meta path importer that uses a dict for back-end storage for source code. While not an optimal solution – manipulations of __path__ on packages does not influence import – it does illustrate what little is required to implement an importer.

"""An importer where source is stored in a dict."""
from importlib import abc


class DictImporter(abc.Finder, abc.PyLoader):

    """A meta path importer that stores source code in a dict.

    The keys are the module names -- packages must end in ``.__init__``.
    The values must be something that can be passed to 'bytes'.

    """

    def __init__(self, memory):
        """Store the dict."""
        self.memory = memory

    def contains(self, name):
        """See if a module or package is in the dict."""
        if name in self.memory:
            return name
        package_name = '{}.__init__'.format(name)
        if  package_name in self.memory:
            return package_name
        return False

    __contains__ = contains  # Convenience.

    def find_module(self, fullname, path=None):
        """Find the module in the dict."""
        if fullname in self:
            return self
        return None

    def source_path(self, fullname):
        """Return the module name if the module is in the dict."""
        if not fullname in self:
            raise ImportError
        return fullname

    def get_data(self, path):
        """Return the bytes for the source.

        The value found in the dict is passed through 'bytes' before being
        returned.

        """
        name = self.contains(path)
        if not name:
            raise IOError
        return bytes(self.memory[name])

    def is_package(self, fullname):
        """Tell if module is a package based on whether the dict contains the
        name with ``.__init__`` appended to it."""
        if fullname not in self:
            raise ImportError
        if fullname in self.memory:
            return False
        # If name is in this importer but not as it is then it must end in
        # ``__init__``.
        else:
            return True