22

✏️ Editing: import.html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
 <head>
 <meta http-equiv="X-UA-Compatible" content="IE=Edge" />
 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
 <title>5. The import system &#8212; Python 3.6.7 documentation</title>
 <link rel="stylesheet" href="../_static/pydoctheme.css" type="text/css" />
 <link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
 <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script>
 <script type="text/javascript" src="../_static/jquery.js"></script>
 <script type="text/javascript" src="../_static/underscore.js"></script>
 <script type="text/javascript" src="../_static/doctools.js"></script>
 <script type="text/javascript" src="../_static/sidebar.js"></script>
 <link rel="search" type="application/opensearchdescription+xml"
 title="Search within Python 3.6.7 documentation"
 href="../_static/opensearch.xml"/>
 <link rel="author" title="About these documents" href="../about.html" />
 <link rel="index" title="Index" href="../genindex.html" />
 <link rel="search" title="Search" href="../search.html" />
 <link rel="copyright" title="Copyright" href="../copyright.html" />
 <link rel="next" title="6. Expressions" href="expressions.html" />
 <link rel="prev" title="4. Execution model" href="executionmodel.html" />
 <link rel="shortcut icon" type="image/png" href="../_static/py.png" />
 <link rel="canonical" href="https://docs.python.org/3/reference/import.html" />
 
 <script type="text/javascript" src="../_static/copybutton.js"></script>

</head><body> 
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 accesskey="I">index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="expressions.html" title="6. Expressions"
 accesskey="N">next</a> |</li>
 <li class="right" >
 <a href="executionmodel.html" title="4. Execution model"
 accesskey="P">previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="https://www.python.org/">Python</a> &#187;</li>
 <li>
 <a href="../index.html">3.6.7 Documentation</a> &#187;
 </li>

<li class="nav-item nav-item-1"><a href="index.html" accesskey="U">The Python Language Reference</a> &#187;</li>
 <li class="right">

<div class="inline-search" style="display: none" role="search">
 <form class="inline-search" action="../search.html" method="get">
 <input placeholder="Quick search" type="text" name="q" />
 <input type="submit" value="Go" />
 <input type="hidden" name="check_keywords" value="yes" />
 <input type="hidden" name="area" value="default" />
 </form>
 </div>
 <script type="text/javascript">$('.inline-search').show(0);</script>
 |
 </li>

</ul>
 </div>

<div class="document">
 <div class="documentwrapper">
 <div class="bodywrapper">
 <div class="body" role="main">
 
 <div class="section" id="the-import-system">
<h1>5. The import system<a class="headerlink" href="#the-import-system" title="Permalink to this headline">¶</a></h1>
Python code in one <a class="reference internal" href="../glossary.html#term-module">module</a> gains access to the code in another module
by the process of <a class="reference internal" href="../glossary.html#term-importing">importing</a> it. The <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement is
the most common way of invoking the import machinery, but it is not the only
way. Functions such as <a class="reference internal" href="../library/importlib.html#importlib.import_module" title="importlib.import_module"><code class="xref py py-func docutils literal notranslate">importlib.import_module()</code></a> and built-in
<a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> can also be used to invoke the import machinery.
The <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement combines two operations; it searches for the
named module, then it binds the results of that search to a name in the local
scope. The search operation of the <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement is defined as
a call to the <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> function, with the appropriate arguments.
The return value of <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> is used to perform the name
binding operation of the <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement. See the
<a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement for the exact details of that name binding
operation.
A direct call to <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> performs only the module search and, if
found, the module creation operation. While certain side-effects may occur,
such as the importing of parent packages, and the updating of various caches
(including <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>), only the <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement performs
a name binding operation.
When calling <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> as part of an import statement, the
standard builtin <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> is called. Other mechanisms for
invoking the import system (such as <a class="reference internal" href="../library/importlib.html#importlib.import_module" title="importlib.import_module"><code class="xref py py-func docutils literal notranslate">importlib.import_module()</code></a>) may
choose to subvert <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> and use its own solution to
implement import semantics.
When a module is first imported, Python searches for the module and if found,
it creates a module object <a class="footnote-reference" href="#fnmo" id="id1">[1]</a>, initializing it. If the named module
cannot be found, a <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a> is raised. Python implements various
strategies to search for the named module when the import machinery is
invoked. These strategies can be modified and extended by using various hooks
described in the sections below.
<div class="versionchanged">
Changed in version 3.3: The import system has been updated to fully implement the second phase
of <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0302">PEP 302</a>. There is no longer any implicit import machinery - the full
import system is exposed through <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a>. In addition,
native namespace package support has been implemented (see <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a>).
</div>
<div class="section" id="importlib">
<h2>5.1. <a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a><a class="headerlink" href="#importlib" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> module provides a rich API for interacting with the
import system. For example <a class="reference internal" href="../library/importlib.html#importlib.import_module" title="importlib.import_module"><code class="xref py py-func docutils literal notranslate">importlib.import_module()</code></a> provides a
recommended, simpler API than built-in <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> for invoking the
import machinery. Refer to the <a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> library documentation for
additional detail.
</div>
<div class="section" id="packages">
<h2>5.2. Packages<a class="headerlink" href="#packages" title="Permalink to this headline">¶</a></h2>
Python has only one type of module object, and all modules are of this type,
regardless of whether the module is implemented in Python, C, or something
else. To help organize modules and provide a naming hierarchy, Python has a
concept of <a class="reference internal" href="../glossary.html#term-package">packages</a>.
You can think of packages as the directories on a file system and modules as
files within directories, but don’t take this analogy too literally since
packages and modules need not originate from the file system. For the
purposes of this documentation, we’ll use this convenient analogy of
directories and files. Like file system directories, packages are organized
hierarchically, and packages may themselves contain subpackages, as well as
regular modules.
It’s important to keep in mind that all packages are modules, but not all
modules are packages. Or put another way, packages are just a special kind of
module. Specifically, any module that contains a <code class="docutils literal notranslate">__path__</code> attribute is
considered a package.
All modules have a name. Subpackage names are separated from their parent
package name by dots, akin to Python’s standard attribute access syntax. Thus
you might have a module called <a class="reference internal" href="../library/sys.html#module-sys" title="sys: Access system-specific parameters and functions."><code class="xref py py-mod docutils literal notranslate">sys</code></a> and a package called <a class="reference internal" href="../library/email.html#module-email" title="email: Package supporting the parsing, manipulating, and generating email messages."><code class="xref py py-mod docutils literal notranslate">email</code></a>,
which in turn has a subpackage called <a class="reference internal" href="../library/email.mime.html#module-email.mime" title="email.mime: Build MIME messages."><code class="xref py py-mod docutils literal notranslate">email.mime</code></a> and a module within
that subpackage called <code class="xref py py-mod docutils literal notranslate">email.mime.text</code>.
<div class="section" id="regular-packages">
<h3>5.2.1. Regular packages<a class="headerlink" href="#regular-packages" title="Permalink to this headline">¶</a></h3>
Python defines two types of packages, <a class="reference internal" href="../glossary.html#term-regular-package">regular packages</a> and <a class="reference internal" href="../glossary.html#term-namespace-package">namespace packages</a>. Regular
packages are traditional packages as they existed in Python 3.2 and earlier.
A regular package is typically implemented as a directory containing an
<code class="docutils literal notranslate">__init__.py</code> file. When a regular package is imported, this
<code class="docutils literal notranslate">__init__.py</code> file is implicitly executed, and the objects it defines are
bound to names in the package’s namespace. The <code class="docutils literal notranslate">__init__.py</code> file can
contain the same Python code that any other module can contain, and Python
will add some additional attributes to the module when it is imported.
For example, the following file system layout defines a top level <code class="docutils literal notranslate">parent</code>
package with three subpackages:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>parent/
 __init__.py
 one/
 __init__.py
 two/
 __init__.py
 three/
 __init__.py
</pre></div>
</div>
Importing <code class="docutils literal notranslate">parent.one</code> will implicitly execute <code class="docutils literal notranslate">parent/__init__.py</code> and
<code class="docutils literal notranslate">parent/one/__init__.py</code>. Subsequent imports of <code class="docutils literal notranslate">parent.two</code> or
<code class="docutils literal notranslate">parent.three</code> will execute <code class="docutils literal notranslate">parent/two/__init__.py</code> and
<code class="docutils literal notranslate">parent/three/__init__.py</code> respectively.
</div>
<div class="section" id="namespace-packages">
<h3>5.2.2. Namespace packages<a class="headerlink" href="#namespace-packages" title="Permalink to this headline">¶</a></h3>
A namespace package is a composite of various <a class="reference internal" href="../glossary.html#term-portion">portions</a>,
where each portion contributes a subpackage to the parent package. Portions
may reside in different locations on the file system. Portions may also be
found in zip files, on the network, or anywhere else that Python searches
during import. Namespace packages may or may not correspond directly to
objects on the file system; they may be virtual modules that have no concrete
representation.
Namespace packages do not use an ordinary list for their <code class="docutils literal notranslate">__path__</code>
attribute. They instead use a custom iterable type which will automatically
perform a new search for package portions on the next import attempt within
that package if the path of their parent package (or <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> for a
top level package) changes.
With namespace packages, there is no <code class="docutils literal notranslate">parent/__init__.py</code> file. In fact,
there may be multiple <code class="docutils literal notranslate">parent</code> directories found during import search, where
each one is provided by a different portion. Thus <code class="docutils literal notranslate">parent/one</code> may not be
physically located next to <code class="docutils literal notranslate">parent/two</code>. In this case, Python will create a
namespace package for the top-level <code class="docutils literal notranslate">parent</code> package whenever it or one of
its subpackages is imported.
See also <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a> for the namespace package specification.
</div>
</div>
<div class="section" id="searching">
<h2>5.3. Searching<a class="headerlink" href="#searching" title="Permalink to this headline">¶</a></h2>
To begin the search, Python needs the <a class="reference internal" href="../glossary.html#term-qualified-name">fully qualified</a>
name of the module (or package, but for the purposes of this discussion, the
difference is immaterial) being imported. This name may come from various
arguments to the <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement, or from the parameters to the
<a class="reference internal" href="../library/importlib.html#importlib.import_module" title="importlib.import_module"><code class="xref py py-func docutils literal notranslate">importlib.import_module()</code></a> or <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> functions.
This name will be used in various phases of the import search, and it may be
the dotted path to a submodule, e.g. <code class="docutils literal notranslate">foo.bar.baz</code>. In this case, Python
first tries to import <code class="docutils literal notranslate">foo</code>, then <code class="docutils literal notranslate">foo.bar</code>, and finally <code class="docutils literal notranslate">foo.bar.baz</code>.
If any of the intermediate imports fail, a <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a> is raised.
<div class="section" id="the-module-cache">
<h3>5.3.1. The module cache<a class="headerlink" href="#the-module-cache" title="Permalink to this headline">¶</a></h3>
The first place checked during import search is <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>. This
mapping serves as a cache of all modules that have been previously imported,
including the intermediate paths. So if <code class="docutils literal notranslate">foo.bar.baz</code> was previously
imported, <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> will contain entries for <code class="docutils literal notranslate">foo</code>, <code class="docutils literal notranslate">foo.bar</code>,
and <code class="docutils literal notranslate">foo.bar.baz</code>. Each key will have as its value the corresponding module
object.
During import, the module name is looked up in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> and if
present, the associated value is the module satisfying the import, and the
process completes. However, if the value is <code class="docutils literal notranslate">None</code>, then a
<a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a> is raised. If the module name is missing, Python will
continue searching for the module.
<a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> is writable. Deleting a key may not destroy the
associated module (as other modules may hold references to it),
but it will invalidate the cache entry for the named module, causing
Python to search anew for the named module upon its next
import. The key can also be assigned to <code class="docutils literal notranslate">None</code>, forcing the next import
of the module to result in a <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a>.
Beware though, as if you keep a reference to the module object,
invalidate its cache entry in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, and then re-import the
named module, the two module objects will not be the same. By contrast,
<a class="reference internal" href="../library/importlib.html#importlib.reload" title="importlib.reload"><code class="xref py py-func docutils literal notranslate">importlib.reload()</code></a> will reuse the same module object, and simply
reinitialise the module contents by rerunning the module’s code.
</div>
<div class="section" id="finders-and-loaders">
<h3>5.3.2. Finders and loaders<a class="headerlink" href="#finders-and-loaders" title="Permalink to this headline">¶</a></h3>
If the named module is not found in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, then Python’s import
protocol is invoked to find and load the module. This protocol consists of
two conceptual objects, <a class="reference internal" href="../glossary.html#term-finder">finders</a> and <a class="reference internal" href="../glossary.html#term-loader">loaders</a>.
A finder’s job is to determine whether it can find the named module using
whatever strategy it knows about. Objects that implement both of these
interfaces are referred to as <a class="reference internal" href="../glossary.html#term-importer">importers</a> - they return
themselves when they find that they can load the requested module.
Python includes a number of default finders and importers. The first one
knows how to locate built-in modules, and the second knows how to locate
frozen modules. A third default finder searches an <a class="reference internal" href="../glossary.html#term-import-path">import path</a>
for modules. The <a class="reference internal" href="../glossary.html#term-import-path">import path</a> is a list of locations that may
name file system paths or zip files. It can also be extended to search
for any locatable resource, such as those identified by URLs.
The import machinery is extensible, so new finders can be added to extend the
range and scope of module searching.
Finders do not actually load modules. If they can find the named module, they
return a module spec, an encapsulation of the module’s import-related
information, which the import machinery then uses when loading the module.
The following sections describe the protocol for finders and loaders in more
detail, including how you can create and register new ones to extend the
import machinery.
<div class="versionchanged">
Changed in version 3.4: In previous versions of Python, finders returned <a class="reference internal" href="../glossary.html#term-loader">loaders</a>
directly, whereas now they return module specs which contain loaders.
Loaders are still used during import but have fewer responsibilities.
</div>
</div>
<div class="section" id="import-hooks">
<h3>5.3.3. Import hooks<a class="headerlink" href="#import-hooks" title="Permalink to this headline">¶</a></h3>
The import machinery is designed to be extensible; the primary mechanism for
this are the import hooks. There are two types of import hooks: meta
hooks and import path hooks.
Meta hooks are called at the start of import processing, before any other
import processing has occurred, other than <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> cache look up.
This allows meta hooks to override <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> processing, frozen
modules, or even built-in modules. Meta hooks are registered by adding new
finder objects to <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a>, as described below.
Import path hooks are called as part of <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> (or
<code class="docutils literal notranslate">package.__path__</code>) processing, at the point where their associated path
item is encountered. Import path hooks are registered by adding new callables
to <a class="reference internal" href="../library/sys.html#sys.path_hooks" title="sys.path_hooks"><code class="xref py py-data docutils literal notranslate">sys.path_hooks</code></a> as described below.
</div>
<div class="section" id="the-meta-path">
<h3>5.3.4. The meta path<a class="headerlink" href="#the-meta-path" title="Permalink to this headline">¶</a></h3>
When the named module is not found in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, Python next
searches <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a>, which contains a list of meta path finder
objects. These finders are queried in order to see if they know how to handle
the named module. Meta path finders must implement a method called
<a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_spec" title="importlib.abc.MetaPathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> which takes three arguments:
a name, an import path, and (optionally) a target module. The meta path
finder can use any strategy it wants to determine whether it can handle
the named module or not.
If the meta path finder knows how to handle the named module, it returns a
spec object. If it cannot handle the named module, it returns <code class="docutils literal notranslate">None</code>. If
<a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a> processing reaches the end of its list without returning
a spec, then a <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a> is raised. Any other exceptions
raised are simply propagated up, aborting the import process.
The <a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_spec" title="importlib.abc.MetaPathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> method of meta path
finders is called with two or three arguments. The first is the fully
qualified name of the module being imported, for example <code class="docutils literal notranslate">foo.bar.baz</code>.
The second argument is the path entries to use for the module search. For
top-level modules, the second argument is <code class="docutils literal notranslate">None</code>, but for submodules or
subpackages, the second argument is the value of the parent package’s
<code class="docutils literal notranslate">__path__</code> attribute. If the appropriate <code class="docutils literal notranslate">__path__</code> attribute cannot
be accessed, a <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a> is raised. The third argument
is an existing module object that will be the target of loading later.
The import system passes in a target module only during reload.
The meta path may be traversed multiple times for a single import request.
For example, assuming none of the modules involved has already been cached,
importing <code class="docutils literal notranslate">foo.bar.baz</code> will first perform a top level import, calling
<code class="docutils literal notranslate">mpf.find_spec(&quot;foo&quot;, None, None)</code> on each meta path finder (<code class="docutils literal notranslate">mpf</code>). After
<code class="docutils literal notranslate">foo</code> has been imported, <code class="docutils literal notranslate">foo.bar</code> will be imported by traversing the
meta path a second time, calling
<code class="docutils literal notranslate">mpf.find_spec(&quot;foo.bar&quot;, foo.__path__, None)</code>. Once <code class="docutils literal notranslate">foo.bar</code> has been
imported, the final traversal will call
<code class="docutils literal notranslate">mpf.find_spec(&quot;foo.bar.baz&quot;, foo.bar.__path__, None)</code>.
Some meta path finders only support top level imports. These importers will
always return <code class="docutils literal notranslate">None</code> when anything other than <code class="docutils literal notranslate">None</code> is passed as the
second argument.
Python’s default <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a> has three meta path finders, one that
knows how to import built-in modules, one that knows how to import frozen
modules, and one that knows how to import modules from an <a class="reference internal" href="../glossary.html#term-import-path">import path</a>
(i.e. the <a class="reference internal" href="../glossary.html#term-path-based-finder">path based finder</a>).
<div class="versionchanged">
Changed in version 3.4: The <a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_spec" title="importlib.abc.MetaPathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> method of meta path
finders replaced <a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_module" title="importlib.abc.MetaPathFinder.find_module"><code class="xref py py-meth docutils literal notranslate">find_module()</code></a>, which
is now deprecated. While it will continue to work without change, the
import machinery will try it only if the finder does not implement
<code class="docutils literal notranslate">find_spec()</code>.
</div>
</div>
</div>
<div class="section" id="loading">
<h2>5.4. Loading<a class="headerlink" href="#loading" title="Permalink to this headline">¶</a></h2>
If and when a module spec is found, the import machinery will use it (and
the loader it contains) when loading the module. Here is an approximation
of what happens during the loading portion of import:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>module = None
if spec.loader is not None and hasattr(spec.loader, &#39;create_module&#39;):
 # It is assumed &#39;exec_module&#39; will also be defined on the loader.
 module = spec.loader.create_module(spec)
if module is None:
 module = ModuleType(spec.name)
# The import-related module attributes get set here:
_init_module_attrs(spec, module)

if spec.loader is None:
 if spec.submodule_search_locations is not None:
 # namespace package
 sys.modules[spec.name] = module
 else:
 # unsupported
 raise ImportError
elif not hasattr(spec.loader, &#39;exec_module&#39;):
 module = spec.loader.load_module(spec.name)
 # Set __loader__ and __package__ if missing.
else:
 sys.modules[spec.name] = module
 try:
 spec.loader.exec_module(module)
 except BaseException:
 try:
 del sys.modules[spec.name]
 except KeyError:
 pass
 raise
return sys.modules[spec.name]
</pre></div>
</div>
Note the following details:
<blockquote>
<div><ul class="simple">
<li>If there is an existing module object with the given name in
<a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, import will have already returned it.</li>
<li>The module will exist in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> before the loader
executes the module code. This is crucial because the module code may
(directly or indirectly) import itself; adding it to <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>
beforehand prevents unbounded recursion in the worst case and multiple
loading in the best.</li>
<li>If loading fails, the failing module – and only the failing module –
gets removed from <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>. Any module already in the
<a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> cache, and any module that was successfully loaded
as a side-effect, must remain in the cache. This contrasts with
reloading where even the failing module is left in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>.</li>
<li>After the module is created but before execution, the import machinery
sets the import-related module attributes (“_init_module_attrs” in
the pseudo-code example above), as summarized in a
<a class="reference internal" href="#import-mod-attrs">later section</a>.</li>
<li>Module execution is the key moment of loading in which the module’s
namespace gets populated. Execution is entirely delegated to the
loader, which gets to decide what gets populated and how.</li>
<li>The module created during loading and passed to exec_module() may
not be the one returned at the end of import <a class="footnote-reference" href="#fnlo" id="id2">[2]</a>.</li>
</ul>
</div></blockquote>
<div class="versionchanged">
Changed in version 3.4: The import system has taken over the boilerplate responsibilities of
loaders. These were previously performed by the
<a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.load_module" title="importlib.abc.Loader.load_module"><code class="xref py py-meth docutils literal notranslate">importlib.abc.Loader.load_module()</code></a> method.
</div>
<div class="section" id="loaders">
<h3>5.4.1. Loaders<a class="headerlink" href="#loaders" title="Permalink to this headline">¶</a></h3>
Module loaders provide the critical function of loading: module execution.
The import machinery calls the <a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.exec_module" title="importlib.abc.Loader.exec_module"><code class="xref py py-meth docutils literal notranslate">importlib.abc.Loader.exec_module()</code></a>
method with a single argument, the module object to execute. Any value
returned from <a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.exec_module" title="importlib.abc.Loader.exec_module"><code class="xref py py-meth docutils literal notranslate">exec_module()</code></a> is ignored.
Loaders must satisfy the following requirements:
<blockquote>
<div><ul class="simple">
<li>If the module is a Python module (as opposed to a built-in module or a
dynamically loaded extension), the loader should execute the module’s code
in the module’s global name space (<code class="docutils literal notranslate">module.__dict__</code>).</li>
<li>If the loader cannot execute the module, it should raise an
<a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a>, although any other exception raised during
<a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.exec_module" title="importlib.abc.Loader.exec_module"><code class="xref py py-meth docutils literal notranslate">exec_module()</code></a> will be propagated.</li>
</ul>
</div></blockquote>
In many cases, the finder and loader can be the same object; in such cases the
<a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_spec" title="importlib.abc.MetaPathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> method would just return a
spec with the loader set to <code class="docutils literal notranslate">self</code>.
Module loaders may opt in to creating the module object during loading
by implementing a <a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.create_module" title="importlib.abc.Loader.create_module"><code class="xref py py-meth docutils literal notranslate">create_module()</code></a> method.
It takes one argument, the module spec, and returns the new module object
to use during loading. <code class="docutils literal notranslate">create_module()</code> does not need to set any attributes
on the module object. If the method returns <code class="docutils literal notranslate">None</code>, the
import machinery will create the new module itself.
<div class="versionadded">
New in version 3.4: The <a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.create_module" title="importlib.abc.Loader.create_module"><code class="xref py py-meth docutils literal notranslate">create_module()</code></a> method of loaders.
</div>
<div class="versionchanged">
Changed in version 3.4: The <a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.load_module" title="importlib.abc.Loader.load_module"><code class="xref py py-meth docutils literal notranslate">load_module()</code></a> method was replaced by
<a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.exec_module" title="importlib.abc.Loader.exec_module"><code class="xref py py-meth docutils literal notranslate">exec_module()</code></a> and the import
machinery assumed all the boilerplate responsibilities of loading.
For compatibility with existing loaders, the import machinery will use
the <code class="docutils literal notranslate">load_module()</code> method of loaders if it exists and the loader does
not also implement <code class="docutils literal notranslate">exec_module()</code>. However, <code class="docutils literal notranslate">load_module()</code> has been
deprecated and loaders should implement <code class="docutils literal notranslate">exec_module()</code> instead.
The <code class="docutils literal notranslate">load_module()</code> method must implement all the boilerplate loading
functionality described above in addition to executing the module. All
the same constraints apply, with some additional clarification:
<blockquote>
<div><ul class="simple">
<li>If there is an existing module object with the given name in
<a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, the loader must use that existing module.
(Otherwise, <a class="reference internal" href="../library/importlib.html#importlib.reload" title="importlib.reload"><code class="xref py py-func docutils literal notranslate">importlib.reload()</code></a> will not work correctly.) If the
named module does not exist in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, the loader
must create a new module object and add it to <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>.</li>
<li>The module must exist in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a> before the loader
executes the module code, to prevent unbounded recursion or multiple
loading.</li>
<li>If loading fails, the loader must remove any modules it has inserted
into <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>, but it must remove only the failing
module(s), and only if the loader itself has loaded the module(s)
explicitly.</li>
</ul>
</div></blockquote>
</div>
<div class="versionchanged">
Changed in version 3.5: A <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate">DeprecationWarning</code></a> is raised when <code class="docutils literal notranslate">exec_module()</code> is defined but
<code class="docutils literal notranslate">create_module()</code> is not.
</div>
<div class="versionchanged">
Changed in version 3.6: An <a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a> is raised when <code class="docutils literal notranslate">exec_module()</code> is defined but
<code class="docutils literal notranslate">create_module()</code> is not.
</div>
</div>
<div class="section" id="submodules">
<h3>5.4.2. Submodules<a class="headerlink" href="#submodules" title="Permalink to this headline">¶</a></h3>
When a submodule is loaded using any mechanism (e.g. <code class="docutils literal notranslate">importlib</code> APIs, the
<code class="docutils literal notranslate">import</code> or <code class="docutils literal notranslate">import-from</code> statements, or built-in <code class="docutils literal notranslate">__import__()</code>) a
binding is placed in the parent module’s namespace to the submodule object.
For example, if package <code class="docutils literal notranslate">spam</code> has a submodule <code class="docutils literal notranslate">foo</code>, after importing
<code class="docutils literal notranslate">spam.foo</code>, <code class="docutils literal notranslate">spam</code> will have an attribute <code class="docutils literal notranslate">foo</code> which is bound to the
submodule. Let’s say you have the following directory structure:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>spam/
 __init__.py
 foo.py
 bar.py
</pre></div>
</div>
and <code class="docutils literal notranslate">spam/__init__.py</code> has the following lines in it:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>from .foo import Foo
from .bar import Bar
</pre></div>
</div>
then executing the following puts a name binding to <code class="docutils literal notranslate">foo</code> and <code class="docutils literal notranslate">bar</code> in the
<code class="docutils literal notranslate">spam</code> module:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import spam
&gt;&gt;&gt; spam.foo
&lt;module &#39;spam.foo&#39; from &#39;/tmp/imports/spam/foo.py&#39;&gt;
&gt;&gt;&gt; spam.bar
&lt;module &#39;spam.bar&#39; from &#39;/tmp/imports/spam/bar.py&#39;&gt;
</pre></div>
</div>
Given Python’s familiar name binding rules this might seem surprising, but
it’s actually a fundamental feature of the import system. The invariant
holding is that if you have <code class="docutils literal notranslate">sys.modules['spam']</code> and
<code class="docutils literal notranslate">sys.modules['spam.foo']</code> (as you would after the above import), the latter
must appear as the <code class="docutils literal notranslate">foo</code> attribute of the former.
</div>
<div class="section" id="module-spec">
<h3>5.4.3. Module spec<a class="headerlink" href="#module-spec" title="Permalink to this headline">¶</a></h3>
The import machinery uses a variety of information about each module
during import, especially before loading. Most of the information is
common to all modules. The purpose of a module’s spec is to encapsulate
this import-related information on a per-module basis.
Using a spec during import allows state to be transferred between import
system components, e.g. between the finder that creates the module spec
and the loader that executes it. Most importantly, it allows the
import machinery to perform the boilerplate operations of loading,
whereas without a module spec the loader had that responsibility.
The module’s spec is exposed as the <code class="docutils literal notranslate">__spec__</code> attribute on a module object.
See <a class="reference internal" href="../library/importlib.html#importlib.machinery.ModuleSpec" title="importlib.machinery.ModuleSpec"><code class="xref py py-class docutils literal notranslate">ModuleSpec</code></a> for details on the contents of
the module spec.
<div class="versionadded">
New in version 3.4.
</div>
</div>
<div class="section" id="import-related-module-attributes">
<h3>5.4.4. Import-related module attributes<a class="headerlink" href="#import-related-module-attributes" title="Permalink to this headline">¶</a></h3>
The import machinery fills in these attributes on each module object
during loading, based on the module’s spec, before the loader executes
the module.
<dl class="attribute">
<dt id="__name__">
<code class="descname">__name__</code><a class="headerlink" href="#__name__" title="Permalink to this definition">¶</a></dt>
<dd>The <code class="docutils literal notranslate">__name__</code> attribute must be set to the fully-qualified name of
the module. This name is used to uniquely identify the module in
the import system.
</dd></dl>

<dl class="attribute">
<dt id="__loader__">
<code class="descname">__loader__</code><a class="headerlink" href="#__loader__" title="Permalink to this definition">¶</a></dt>
<dd>The <code class="docutils literal notranslate">__loader__</code> attribute must be set to the loader object that
the import machinery used when loading the module. This is mostly
for introspection, but can be used for additional loader-specific
functionality, for example getting data associated with a loader.
</dd></dl>

<dl class="attribute">
<dt id="__package__">
<code class="descname">__package__</code><a class="headerlink" href="#__package__" title="Permalink to this definition">¶</a></dt>
<dd>The module’s <code class="docutils literal notranslate">__package__</code> attribute must be set. Its value must
be a string, but it can be the same value as its <code class="docutils literal notranslate">__name__</code>. When
the module is a package, its <code class="docutils literal notranslate">__package__</code> value should be set to
its <code class="docutils literal notranslate">__name__</code>. When the module is not a package, <code class="docutils literal notranslate">__package__</code>
should be set to the empty string for top-level modules, or for
submodules, to the parent package’s name. See <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366">PEP 366</a> for further
details.
This attribute is used instead of <code class="docutils literal notranslate">__name__</code> to calculate explicit
relative imports for main modules, as defined in <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366">PEP 366</a>. It is
expected to have the same value as <code class="docutils literal notranslate">__spec__.parent</code>.
<div class="versionchanged">
Changed in version 3.6: The value of <code class="docutils literal notranslate">__package__</code> is expected to be the same as
<code class="docutils literal notranslate">__spec__.parent</code>.
</div>
</dd></dl>

<dl class="attribute">
<dt id="__spec__">
<code class="descname">__spec__</code><a class="headerlink" href="#__spec__" title="Permalink to this definition">¶</a></dt>
<dd>The <code class="docutils literal notranslate">__spec__</code> attribute must be set to the module spec that was
used when importing the module. Setting <code class="docutils literal notranslate">__spec__</code>
appropriately applies equally to <a class="reference internal" href="toplevel_components.html#programs">modules initialized during
interpreter startup</a>. The one exception is <code class="docutils literal notranslate">__main__</code>,
where <code class="docutils literal notranslate">__spec__</code> is <a class="reference internal" href="#main-spec">set to None in some cases</a>.
When <code class="docutils literal notranslate">__package__</code> is not defined, <code class="docutils literal notranslate">__spec__.parent</code> is used as
a fallback.
<div class="versionadded">
New in version 3.4.
</div>
<div class="versionchanged">
Changed in version 3.6: <code class="docutils literal notranslate">__spec__.parent</code> is used as a fallback when <code class="docutils literal notranslate">__package__</code> is
not defined.
</div>
</dd></dl>

<dl class="attribute">
<dt id="__path__">
<code class="descname">__path__</code><a class="headerlink" href="#__path__" title="Permalink to this definition">¶</a></dt>
<dd>If the module is a package (either regular or namespace), the module
object’s <code class="docutils literal notranslate">__path__</code> attribute must be set. The value must be
iterable, but may be empty if <code class="docutils literal notranslate">__path__</code> has no further significance.
If <code class="docutils literal notranslate">__path__</code> is not empty, it must produce strings when iterated
over. More details on the semantics of <code class="docutils literal notranslate">__path__</code> are given
<a class="reference internal" href="#package-path-rules">below</a>.
Non-package modules should not have a <code class="docutils literal notranslate">__path__</code> attribute.
</dd></dl>

<dl class="attribute">
<dt id="__cached__">
<code class="descname">__cached__</code><a class="headerlink" href="#__cached__" title="Permalink to this definition">¶</a></dt>
<dd><code class="docutils literal notranslate">__file__</code> is optional. If set, this attribute’s value must be a
string. The import system may opt to leave <code class="docutils literal notranslate">__file__</code> unset if it
has no semantic meaning (e.g. a module loaded from a database).
If <code class="docutils literal notranslate">__file__</code> is set, it may also be appropriate to set the
<code class="docutils literal notranslate">__cached__</code> attribute which is the path to any compiled version of
the code (e.g. byte-compiled file). The file does not need to exist
to set this attribute; the path can simply point to where the
compiled file would exist (see <a class="pep reference external" href="https://www.python.org/dev/peps/pep-3147">PEP 3147</a>).
It is also appropriate to set <code class="docutils literal notranslate">__cached__</code> when <code class="docutils literal notranslate">__file__</code> is not
set. However, that scenario is quite atypical. Ultimately, the
loader is what makes use of <code class="docutils literal notranslate">__file__</code> and/or <code class="docutils literal notranslate">__cached__</code>. So
if a loader can load from a cached module but otherwise does not load
from a file, that atypical scenario may be appropriate.
</dd></dl>

</div>
<div class="section" id="module-path">
<h3>5.4.5. module.__path__<a class="headerlink" href="#module-path" title="Permalink to this headline">¶</a></h3>
By definition, if a module has a <code class="docutils literal notranslate">__path__</code> attribute, it is a package.
A package’s <code class="docutils literal notranslate">__path__</code> attribute is used during imports of its subpackages.
Within the import machinery, it functions much the same as <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a>,
i.e. providing a list of locations to search for modules during import.
However, <code class="docutils literal notranslate">__path__</code> is typically much more constrained than
<a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a>.
<code class="docutils literal notranslate">__path__</code> must be an iterable of strings, but it may be empty.
The same rules used for <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> also apply to a package’s
<code class="docutils literal notranslate">__path__</code>, and <a class="reference internal" href="../library/sys.html#sys.path_hooks" title="sys.path_hooks"><code class="xref py py-data docutils literal notranslate">sys.path_hooks</code></a> (described below) are
consulted when traversing a package’s <code class="docutils literal notranslate">__path__</code>.
A package’s <code class="docutils literal notranslate">__init__.py</code> file may set or alter the package’s <code class="docutils literal notranslate">__path__</code>
attribute, and this was typically the way namespace packages were implemented
prior to <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a>. With the adoption of <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a>, namespace packages no
longer need to supply <code class="docutils literal notranslate">__init__.py</code> files containing only <code class="docutils literal notranslate">__path__</code>
manipulation code; the import machinery automatically sets <code class="docutils literal notranslate">__path__</code>
correctly for the namespace package.
</div>
<div class="section" id="module-reprs">
<h3>5.4.6. Module reprs<a class="headerlink" href="#module-reprs" title="Permalink to this headline">¶</a></h3>
By default, all modules have a usable repr, however depending on the
attributes set above, and in the module’s spec, you can more explicitly
control the repr of module objects.
If the module has a spec (<code class="docutils literal notranslate">__spec__</code>), the import machinery will try
to generate a repr from it. If that fails or there is no spec, the import
system will craft a default repr using whatever information is available
on the module. It will try to use the <code class="docutils literal notranslate">module.__name__</code>,
<code class="docutils literal notranslate">module.__file__</code>, and <code class="docutils literal notranslate">module.__loader__</code> as input into the repr,
with defaults for whatever information is missing.
Here are the exact rules used:
<blockquote>
<div><ul class="simple">
<li>If the module has a <code class="docutils literal notranslate">__spec__</code> attribute, the information in the spec
is used to generate the repr. The “name”, “loader”, “origin”, and
“has_location” attributes are consulted.</li>
<li>If the module has a <code class="docutils literal notranslate">__file__</code> attribute, this is used as part of the
module’s repr.</li>
<li>If the module has no <code class="docutils literal notranslate">__file__</code> but does have a <code class="docutils literal notranslate">__loader__</code> that is not
<code class="docutils literal notranslate">None</code>, then the loader’s repr is used as part of the module’s repr.</li>
<li>Otherwise, just use the module’s <code class="docutils literal notranslate">__name__</code> in the repr.</li>
</ul>
</div></blockquote>
<div class="versionchanged">
Changed in version 3.4: Use of <a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.module_repr" title="importlib.abc.Loader.module_repr"><code class="xref py py-meth docutils literal notranslate">loader.module_repr()</code></a>
has been deprecated and the module spec is now used by the import
machinery to generate a module repr.
For backward compatibility with Python 3.3, the module repr will be
generated by calling the loader’s
<a class="reference internal" href="../library/importlib.html#importlib.abc.Loader.module_repr" title="importlib.abc.Loader.module_repr"><code class="xref py py-meth docutils literal notranslate">module_repr()</code></a> method, if defined, before
trying either approach described above. However, the method is deprecated.
</div>
</div>
</div>
<div class="section" id="the-path-based-finder">
<h2>5.5. The Path Based Finder<a class="headerlink" href="#the-path-based-finder" title="Permalink to this headline">¶</a></h2>
As mentioned previously, Python comes with several default meta path finders.
One of these, called the <a class="reference internal" href="../glossary.html#term-path-based-finder">path based finder</a>
(<a class="reference internal" href="../library/importlib.html#importlib.machinery.PathFinder" title="importlib.machinery.PathFinder"><code class="xref py py-class docutils literal notranslate">PathFinder</code></a>), searches an <a class="reference internal" href="../glossary.html#term-import-path">import path</a>,
which contains a list of <a class="reference internal" href="../glossary.html#term-path-entry">path entries</a>. Each path
entry names a location to search for modules.
The path based finder itself doesn’t know how to import anything. Instead, it
traverses the individual path entries, associating each of them with a
path entry finder that knows how to handle that particular kind of path.
The default set of path entry finders implement all the semantics for finding
modules on the file system, handling special file types such as Python source
code (<code class="docutils literal notranslate">.py</code> files), Python byte code (<code class="docutils literal notranslate">.pyc</code> files) and
shared libraries (e.g. <code class="docutils literal notranslate">.so</code> files). When supported by the <a class="reference internal" href="../library/zipimport.html#module-zipimport" title="zipimport: support for importing Python modules from ZIP archives."><code class="xref py py-mod docutils literal notranslate">zipimport</code></a>
module in the standard library, the default path entry finders also handle
loading all of these file types (other than shared libraries) from zipfiles.
Path entries need not be limited to file system locations. They can refer to
URLs, database queries, or any other location that can be specified as a
string.
The path based finder provides additional hooks and protocols so that you
can extend and customize the types of searchable path entries. For example,
if you wanted to support path entries as network URLs, you could write a hook
that implements HTTP semantics to find modules on the web. This hook (a
callable) would return a <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a> supporting the protocol
described below, which was then used to get a loader for the module from the
web.
A word of warning: this section and the previous both use the term finder,
distinguishing between them by using the terms <a class="reference internal" href="../glossary.html#term-meta-path-finder">meta path finder</a> and
<a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a>. These two types of finders are very similar,
support similar protocols, and function in similar ways during the import
process, but it’s important to keep in mind that they are subtly different.
In particular, meta path finders operate at the beginning of the import
process, as keyed off the <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a> traversal.
By contrast, path entry finders are in a sense an implementation detail
of the path based finder, and in fact, if the path based finder were to be
removed from <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a>, none of the path entry finder semantics
would be invoked.
<div class="section" id="path-entry-finders">
<h3>5.5.1. Path entry finders<a class="headerlink" href="#path-entry-finders" title="Permalink to this headline">¶</a></h3>
The <a class="reference internal" href="../glossary.html#term-path-based-finder">path based finder</a> is responsible for finding and loading
Python modules and packages whose location is specified with a string
<a class="reference internal" href="../glossary.html#term-path-entry">path entry</a>. Most path entries name locations in the file system,
but they need not be limited to this.
As a meta path finder, the <a class="reference internal" href="../glossary.html#term-path-based-finder">path based finder</a> implements the
<a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_spec" title="importlib.abc.MetaPathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> protocol previously
described, however it exposes additional hooks that can be used to
customize how modules are found and loaded from the <a class="reference internal" href="../glossary.html#term-import-path">import path</a>.
Three variables are used by the <a class="reference internal" href="../glossary.html#term-path-based-finder">path based finder</a>, <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a>,
<a class="reference internal" href="../library/sys.html#sys.path_hooks" title="sys.path_hooks"><code class="xref py py-data docutils literal notranslate">sys.path_hooks</code></a> and <a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a>. The <code class="docutils literal notranslate">__path__</code>
attributes on package objects are also used. These provide additional ways
that the import machinery can be customized.
<a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> contains a list of strings providing search locations for
modules and packages. It is initialized from the <code class="xref py py-data docutils literal notranslate">PYTHONPATH</code>
environment variable and various other installation- and
implementation-specific defaults. Entries in <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> can name
directories on the file system, zip files, and potentially other “locations”
(see the <a class="reference internal" href="../library/site.html#module-site" title="site: Module responsible for site-specific configuration."><code class="xref py py-mod docutils literal notranslate">site</code></a> module) that should be searched for modules, such as
URLs, or database queries. Only strings and bytes should be present on
<a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a>; all other data types are ignored. The encoding of bytes
entries is determined by the individual <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finders</a>.
The <a class="reference internal" href="../glossary.html#term-path-based-finder">path based finder</a> is a <a class="reference internal" href="../glossary.html#term-meta-path-finder">meta path finder</a>, so the import
machinery begins the <a class="reference internal" href="../glossary.html#term-import-path">import path</a> search by calling the path
based finder’s <a class="reference internal" href="../library/importlib.html#importlib.machinery.PathFinder.find_spec" title="importlib.machinery.PathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> method as
described previously. When the <code class="docutils literal notranslate">path</code> argument to
<a class="reference internal" href="../library/importlib.html#importlib.machinery.PathFinder.find_spec" title="importlib.machinery.PathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> is given, it will be a
list of string paths to traverse - typically a package’s <code class="docutils literal notranslate">__path__</code>
attribute for an import within that package. If the <code class="docutils literal notranslate">path</code> argument is
<code class="docutils literal notranslate">None</code>, this indicates a top level import and <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> is used.
The path based finder iterates over every entry in the search path, and
for each of these, looks for an appropriate <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a>
(<a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder" title="importlib.abc.PathEntryFinder"><code class="xref py py-class docutils literal notranslate">PathEntryFinder</code></a>) for the
path entry. Because this can be an expensive operation (e.g. there may be
<cite>stat()</cite> call overheads for this search), the path based finder maintains
a cache mapping path entries to path entry finders. This cache is maintained
in <a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a> (despite the name, this cache actually
stores finder objects rather than being limited to <a class="reference internal" href="../glossary.html#term-importer">importer</a> objects).
In this way, the expensive search for a particular <a class="reference internal" href="../glossary.html#term-path-entry">path entry</a>
location’s <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a> need only be done once. User code is
free to remove cache entries from <a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a> forcing
the path based finder to perform the path entry search again <a class="footnote-reference" href="#fnpic" id="id3">[3]</a>.
If the path entry is not present in the cache, the path based finder iterates
over every callable in <a class="reference internal" href="../library/sys.html#sys.path_hooks" title="sys.path_hooks"><code class="xref py py-data docutils literal notranslate">sys.path_hooks</code></a>. Each of the <a class="reference internal" href="../glossary.html#term-path-entry-hook">path entry
hooks</a> in this list is called with a single argument, the
path entry to be searched. This callable may either return a <a class="reference internal" href="../glossary.html#term-path-entry-finder">path
entry finder</a> that can handle the path entry, or it may raise
<a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a>. An <a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a> is used by the path based finder to
signal that the hook cannot find a <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a>
for that <a class="reference internal" href="../glossary.html#term-path-entry">path entry</a>. The
exception is ignored and <a class="reference internal" href="../glossary.html#term-import-path">import path</a> iteration continues. The hook
should expect either a string or bytes object; the encoding of bytes objects
is up to the hook (e.g. it may be a file system encoding, UTF-8, or something
else), and if the hook cannot decode the argument, it should raise
<a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a>.
If <a class="reference internal" href="../library/sys.html#sys.path_hooks" title="sys.path_hooks"><code class="xref py py-data docutils literal notranslate">sys.path_hooks</code></a> iteration ends with no <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a>
being returned, then the path based finder’s
<a class="reference internal" href="../library/importlib.html#importlib.machinery.PathFinder.find_spec" title="importlib.machinery.PathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> method will store <code class="docutils literal notranslate">None</code>
in <a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a> (to indicate that there is no finder for
this path entry) and return <code class="docutils literal notranslate">None</code>, indicating that this
<a class="reference internal" href="../glossary.html#term-meta-path-finder">meta path finder</a> could not find the module.
If a <a class="reference internal" href="../glossary.html#term-path-entry-finder">path entry finder</a> is returned by one of the <a class="reference internal" href="../glossary.html#term-path-entry-hook">path entry
hook</a> callables on <a class="reference internal" href="../library/sys.html#sys.path_hooks" title="sys.path_hooks"><code class="xref py py-data docutils literal notranslate">sys.path_hooks</code></a>, then the following protocol is used
to ask the finder for a module spec, which is then used when loading the
module.
The current working directory – denoted by an empty string – is handled
slightly differently from other entries on <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a>. First, if the
current working directory is found to not exist, no value is stored in
<a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a>. Second, the value for the current working
directory is looked up fresh for each module lookup. Third, the path used for
<a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a> and returned by
<a class="reference internal" href="../library/importlib.html#importlib.machinery.PathFinder.find_spec" title="importlib.machinery.PathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">importlib.machinery.PathFinder.find_spec()</code></a> will be the actual current
working directory and not the empty string.
</div>
<div class="section" id="path-entry-finder-protocol">
<h3>5.5.2. Path entry finder protocol<a class="headerlink" href="#path-entry-finder-protocol" title="Permalink to this headline">¶</a></h3>
In order to support imports of modules and initialized packages and also to
contribute portions to namespace packages, path entry finders must implement
the <a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder.find_spec" title="importlib.abc.PathEntryFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> method.
<a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder.find_spec" title="importlib.abc.PathEntryFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> takes two argument, the
fully qualified name of the module being imported, and the (optional) target
module. <code class="docutils literal notranslate">find_spec()</code> returns a fully populated spec for the module.
This spec will always have “loader” set (with one exception).
To indicate to the import machinery that the spec represents a namespace
<a class="reference internal" href="../glossary.html#term-portion">portion</a>. the path entry finder sets “loader” on the spec to
<code class="docutils literal notranslate">None</code> and “submodule_search_locations” to a list containing the
portion.
<div class="versionchanged">
Changed in version 3.4: <a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder.find_spec" title="importlib.abc.PathEntryFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> replaced
<a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder.find_loader" title="importlib.abc.PathEntryFinder.find_loader"><code class="xref py py-meth docutils literal notranslate">find_loader()</code></a> and
<a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder.find_module" title="importlib.abc.PathEntryFinder.find_module"><code class="xref py py-meth docutils literal notranslate">find_module()</code></a>, both of which
are now deprecated, but will be used if <code class="docutils literal notranslate">find_spec()</code> is not defined.
Older path entry finders may implement one of these two deprecated methods
instead of <code class="docutils literal notranslate">find_spec()</code>. The methods are still respected for the
sake of backward compatibility. However, if <code class="docutils literal notranslate">find_spec()</code> is
implemented on the path entry finder, the legacy methods are ignored.
<a class="reference internal" href="../library/importlib.html#importlib.abc.PathEntryFinder.find_loader" title="importlib.abc.PathEntryFinder.find_loader"><code class="xref py py-meth docutils literal notranslate">find_loader()</code></a> takes one argument, the
fully qualified name of the module being imported. <code class="docutils literal notranslate">find_loader()</code>
returns a 2-tuple where the first item is the loader and the second item
is a namespace <a class="reference internal" href="../glossary.html#term-portion">portion</a>. When the first item (i.e. the loader) is
<code class="docutils literal notranslate">None</code>, this means that while the path entry finder does not have a
loader for the named module, it knows that the path entry contributes to
a namespace portion for the named module. This will almost always be the
case where Python is asked to import a namespace package that has no
physical presence on the file system. When a path entry finder returns
<code class="docutils literal notranslate">None</code> for the loader, the second item of the 2-tuple return value must
be a sequence, although it can be empty.
If <code class="docutils literal notranslate">find_loader()</code> returns a non-<code class="docutils literal notranslate">None</code> loader value, the portion is
ignored and the loader is returned from the path based finder, terminating
the search through the path entries.
For backwards compatibility with other implementations of the import
protocol, many path entry finders also support the same,
traditional <code class="docutils literal notranslate">find_module()</code> method that meta path finders support.
However path entry finder <code class="docutils literal notranslate">find_module()</code> methods are never called
with a <code class="docutils literal notranslate">path</code> argument (they are expected to record the appropriate
path information from the initial call to the path hook).
The <code class="docutils literal notranslate">find_module()</code> method on path entry finders is deprecated,
as it does not allow the path entry finder to contribute portions to
namespace packages. If both <code class="docutils literal notranslate">find_loader()</code> and <code class="docutils literal notranslate">find_module()</code>
exist on a path entry finder, the import system will always call
<code class="docutils literal notranslate">find_loader()</code> in preference to <code class="docutils literal notranslate">find_module()</code>.
</div>
</div>
</div>
<div class="section" id="replacing-the-standard-import-system">
<h2>5.6. Replacing the standard import system<a class="headerlink" href="#replacing-the-standard-import-system" title="Permalink to this headline">¶</a></h2>
The most reliable mechanism for replacing the entire import system is to
delete the default contents of <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a>, replacing them
entirely with a custom meta path hook.
If it is acceptable to only alter the behaviour of import statements
without affecting other APIs that access the import system, then replacing
the builtin <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate">__import__()</code></a> function may be sufficient. This technique
may also be employed at the module level to only alter the behaviour of
import statements within that module.
To selectively prevent import of some modules from a hook early on the
meta path (rather than disabling the standard import system entirely),
it is sufficient to raise <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate">ModuleNotFoundError</code></a> directly from
<a class="reference internal" href="../library/importlib.html#importlib.abc.MetaPathFinder.find_spec" title="importlib.abc.MetaPathFinder.find_spec"><code class="xref py py-meth docutils literal notranslate">find_spec()</code></a> instead of returning
<code class="docutils literal notranslate">None</code>. The latter indicates that the meta path search should continue,
while raising an exception terminates it immediately.
</div>
<div class="section" id="special-considerations-for-main">
<h2>5.7. Special considerations for __main__<a class="headerlink" href="#special-considerations-for-main" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><code class="xref py py-mod docutils literal notranslate">__main__</code></a> module is a special case relative to Python’s import
system. As noted <a class="reference internal" href="toplevel_components.html#programs">elsewhere</a>, the <code class="docutils literal notranslate">__main__</code> module
is directly initialized at interpreter startup, much like <a class="reference internal" href="../library/sys.html#module-sys" title="sys: Access system-specific parameters and functions."><code class="xref py py-mod docutils literal notranslate">sys</code></a> and
<a class="reference internal" href="../library/builtins.html#module-builtins" title="builtins: The module that provides the built-in namespace."><code class="xref py py-mod docutils literal notranslate">builtins</code></a>. However, unlike those two, it doesn’t strictly
qualify as a built-in module. This is because the manner in which
<code class="docutils literal notranslate">__main__</code> is initialized depends on the flags and other options with
which the interpreter is invoked.
<div class="section" id="main-spec">
<h3>5.7.1. __main__.__spec__<a class="headerlink" href="#main-spec" title="Permalink to this headline">¶</a></h3>
Depending on how <a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><code class="xref py py-mod docutils literal notranslate">__main__</code></a> is initialized, <code class="docutils literal notranslate">__main__.__spec__</code>
gets set appropriately or to <code class="docutils literal notranslate">None</code>.
When Python is started with the <a class="reference internal" href="../using/cmdline.html#cmdoption-m"><code class="xref std std-option docutils literal notranslate">-m</code></a> option, <code class="docutils literal notranslate">__spec__</code> is set
to the module spec of the corresponding module or package. <code class="docutils literal notranslate">__spec__</code> is
also populated when the <code class="docutils literal notranslate">__main__</code> module is loaded as part of executing a
directory, zipfile or other <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate">sys.path</code></a> entry.
In <a class="reference internal" href="../using/cmdline.html#using-on-interface-options">the remaining cases</a>
<code class="docutils literal notranslate">__main__.__spec__</code> is set to <code class="docutils literal notranslate">None</code>, as the code used to populate the
<a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><code class="xref py py-mod docutils literal notranslate">__main__</code></a> does not correspond directly with an importable module:
<ul class="simple">
<li>interactive prompt</li>
<li>-c switch</li>
<li>running from stdin</li>
<li>running directly from a source or bytecode file</li>
</ul>
Note that <code class="docutils literal notranslate">__main__.__spec__</code> is always <code class="docutils literal notranslate">None</code> in the last case,
even if the file could technically be imported directly as a module
instead. Use the <a class="reference internal" href="../using/cmdline.html#cmdoption-m"><code class="xref std std-option docutils literal notranslate">-m</code></a> switch if valid module metadata is desired
in <a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><code class="xref py py-mod docutils literal notranslate">__main__</code></a>.
Note also that even when <code class="docutils literal notranslate">__main__</code> corresponds with an importable module
and <code class="docutils literal notranslate">__main__.__spec__</code> is set accordingly, they’re still considered
distinct modules. This is due to the fact that blocks guarded by
<code class="docutils literal notranslate">if __name__ == &quot;__main__&quot;:</code> checks only execute when the module is used
to populate the <code class="docutils literal notranslate">__main__</code> namespace, and not during normal import.
</div>
</div>
<div class="section" id="open-issues">
<h2>5.8. Open issues<a class="headerlink" href="#open-issues" title="Permalink to this headline">¶</a></h2>
XXX It would be really nice to have a diagram.
XXX * (import_machinery.rst) how about a section devoted just to the
attributes of modules and packages, perhaps expanding upon or supplanting the
related entries in the data model reference page?
XXX runpy, pkgutil, et al in the library manual should all get “See Also”
links at the top pointing to the new import system section.
XXX Add more explanation regarding the different ways in which
<code class="docutils literal notranslate">__main__</code> is initialized?
XXX Add more info on <code class="docutils literal notranslate">__main__</code> quirks/pitfalls (i.e. copy from
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0395">PEP 395</a>).
</div>
<div class="section" id="references">
<h2>5.9. References<a class="headerlink" href="#references" title="Permalink to this headline">¶</a></h2>
The import machinery has evolved considerably since Python’s early days. The
original <a class="reference external" href="https://www.python.org/doc/essays/packages/">specification for packages</a> is still available to read,
although some details have changed since the writing of that document.
The original specification for <a class="reference internal" href="../library/sys.html#sys.meta_path" title="sys.meta_path"><code class="xref py py-data docutils literal notranslate">sys.meta_path</code></a> was <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0302">PEP 302</a>, with
subsequent extension in <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a>.
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a> introduced <a class="reference internal" href="../glossary.html#term-namespace-package">namespace packages</a> for
Python 3.3. <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420">PEP 420</a> also introduced the <code class="xref py py-meth docutils literal notranslate">find_loader()</code> protocol as an
alternative to <code class="xref py py-meth docutils literal notranslate">find_module()</code>.
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366">PEP 366</a> describes the addition of the <code class="docutils literal notranslate">__package__</code> attribute for
explicit relative imports in main modules.
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0328">PEP 328</a> introduced absolute and explicit relative imports and initially
proposed <code class="docutils literal notranslate">__name__</code> for semantics <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366">PEP 366</a> would eventually specify for
<code class="docutils literal notranslate">__package__</code>.
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0338">PEP 338</a> defines executing modules as scripts.
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0451">PEP 451</a> adds the encapsulation of per-module import state in spec
objects. It also off-loads most of the boilerplate responsibilities of
loaders back onto the import machinery. These changes allow the
deprecation of several APIs in the import system and also addition of new
methods to finders and loaders.
Footnotes
<table class="docutils footnote" frame="void" id="fnmo" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id1">[1]</a></td><td>See <a class="reference internal" href="../library/types.html#types.ModuleType" title="types.ModuleType"><code class="xref py py-class docutils literal notranslate">types.ModuleType</code></a>.</td></tr>
</tbody>
</table>
<table class="docutils footnote" frame="void" id="fnlo" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id2">[2]</a></td><td>The importlib implementation avoids using the return value
directly. Instead, it gets the module object by looking the module name up
in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>. The indirect effect of this is that an imported
module may replace itself in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate">sys.modules</code></a>. This is
implementation-specific behavior that is not guaranteed to work in other
Python implementations.</td></tr>
</tbody>
</table>
<table class="docutils footnote" frame="void" id="fnpic" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id3">[3]</a></td><td>In legacy code, it is possible to find instances of
<a class="reference internal" href="../library/imp.html#imp.NullImporter" title="imp.NullImporter"><code class="xref py py-class docutils literal notranslate">imp.NullImporter</code></a> in the <a class="reference internal" href="../library/sys.html#sys.path_importer_cache" title="sys.path_importer_cache"><code class="xref py py-data docutils literal notranslate">sys.path_importer_cache</code></a>. It
is recommended that code be changed to use <code class="docutils literal notranslate">None</code> instead. See
<a class="reference internal" href="../whatsnew/3.3.html#portingpythoncode">Porting Python code</a> for more details.</td></tr>
</tbody>
</table>
</div>
</div>

</div>
 </div>
 </div>
 <div class="sphinxsidebar" role="navigation" aria-label="main navigation">
 <div class="sphinxsidebarwrapper">
 <h3><a href="../contents.html">Table Of Contents</a></h3>
 <ul>
<li><a class="reference internal" href="#">5. The import system</a><ul>
<li><a class="reference internal" href="#importlib">5.1. <code class="docutils literal notranslate">importlib</code></a></li>
<li><a class="reference internal" href="#packages">5.2. Packages</a><ul>
<li><a class="reference internal" href="#regular-packages">5.2.1. Regular packages</a></li>
<li><a class="reference internal" href="#namespace-packages">5.2.2. Namespace packages</a></li>
</ul>
</li>
<li><a class="reference internal" href="#searching">5.3. Searching</a><ul>
<li><a class="reference internal" href="#the-module-cache">5.3.1. The module cache</a></li>
<li><a class="reference internal" href="#finders-and-loaders">5.3.2. Finders and loaders</a></li>
<li><a class="reference internal" href="#import-hooks">5.3.3. Import hooks</a></li>
<li><a class="reference internal" href="#the-meta-path">5.3.4. The meta path</a></li>
</ul>
</li>
<li><a class="reference internal" href="#loading">5.4. Loading</a><ul>
<li><a class="reference internal" href="#loaders">5.4.1. Loaders</a></li>
<li><a class="reference internal" href="#submodules">5.4.2. Submodules</a></li>
<li><a class="reference internal" href="#module-spec">5.4.3. Module spec</a></li>
<li><a class="reference internal" href="#import-related-module-attributes">5.4.4. Import-related module attributes</a></li>
<li><a class="reference internal" href="#module-path">5.4.5. module.__path__</a></li>
<li><a class="reference internal" href="#module-reprs">5.4.6. Module reprs</a></li>
</ul>
</li>
<li><a class="reference internal" href="#the-path-based-finder">5.5. The Path Based Finder</a><ul>
<li><a class="reference internal" href="#path-entry-finders">5.5.1. Path entry finders</a></li>
<li><a class="reference internal" href="#path-entry-finder-protocol">5.5.2. Path entry finder protocol</a></li>
</ul>
</li>
<li><a class="reference internal" href="#replacing-the-standard-import-system">5.6. Replacing the standard import system</a></li>
<li><a class="reference internal" href="#special-considerations-for-main">5.7. Special considerations for __main__</a><ul>
<li><a class="reference internal" href="#main-spec">5.7.1. __main__.__spec__</a></li>
</ul>
</li>
<li><a class="reference internal" href="#open-issues">5.8. Open issues</a></li>
<li><a class="reference internal" href="#references">5.9. References</a></li>
</ul>
</li>
</ul>

<h4>Previous topic</h4>
 <a href="executionmodel.html"
 title="previous chapter">4. Execution model</a>
 <h4>Next topic</h4>
 <a href="expressions.html"
 title="next chapter">6. Expressions</a>
 <div role="note" aria-label="source link">
 <h3>This Page</h3>
 <ul class="this-page-menu">
 <li><a href="../bugs.html">Report a Bug</a></li>
 <li>
 <a href="https://github.com/python/cpython/blob/3.6/Doc/reference/import.rst"
 rel="nofollow">Show Source
 </a>
 </li>
 </ul>
 </div>
 </div>
 </div>
 <div class="clearer"></div>
 </div> 
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 >index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="expressions.html" title="6. Expressions"
 >next</a> |</li>
 <li class="right" >
 <a href="executionmodel.html" title="4. Execution model"
 >previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="https://www.python.org/">Python</a> &#187;</li>
 <li>
 <a href="../index.html">3.6.7 Documentation</a> &#187;
 </li>

<li class="nav-item nav-item-1"><a href="index.html" >The Python Language Reference</a> &#187;</li>
 <li class="right">

</ul>
 </div> 
 <div class="footer">
 &copy; <a href="../copyright.html">Copyright</a> 2001-2023, Python Software Foundation.
 
 The Python Software Foundation is a non-profit corporation.
 <a href="https://www.python.org/psf/donations/">Please donate.</a>
 
 Last updated on Dec 18, 2023.
 <a href="../bugs.html">Found a bug</a>?
 
 Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.7.6.
 </div>

</body>
</html>

MR SKB