D7net
Home
Console
Upload
information
Create File
Create Folder
About
Tools
:
/
usr
/
share
/
doc
/
python3-docs
/
html
/
reference
/
Filename :
import.html
back
Copy
<!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 — 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> »</li> <li> <a href="../index.html">3.6.7 Documentation</a> » </li> <li class="nav-item nav-item-1"><a href="index.html" accesskey="U">The Python Language Reference</a> »</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"> <span id="importsystem"></span><h1>5. The import system<a class="headerlink" href="#the-import-system" title="Permalink to this headline">¶</a></h1> <p id="index-0">Python code in one <a class="reference internal" href="../glossary.html#term-module"><span class="xref std std-term">module</span></a> gains access to the code in another module by the process of <a class="reference internal" href="../glossary.html#term-importing"><span class="xref std std-term">importing</span></a> it. The <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate"><span class="pre">import</span></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"><span class="pre">importlib.import_module()</span></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"><span class="pre">__import__()</span></code></a> can also be used to invoke the import machinery.</p> <p>The <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate"><span class="pre">import</span></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"><span class="pre">import</span></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"><span class="pre">__import__()</span></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"><span class="pre">__import__()</span></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"><span class="pre">import</span></code></a> statement. See the <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate"><span class="pre">import</span></code></a> statement for the exact details of that name binding operation.</p> <p>A direct call to <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate"><span class="pre">__import__()</span></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"><span class="pre">sys.modules</span></code></a>), only the <a class="reference internal" href="simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate"><span class="pre">import</span></code></a> statement performs a name binding operation.</p> <p>When calling <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate"><span class="pre">__import__()</span></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"><span class="pre">__import__()</span></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"><span class="pre">importlib.import_module()</span></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"><span class="pre">__import__()</span></code></a> and use its own solution to implement import semantics.</p> <p>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"><span class="pre">ModuleNotFoundError</span></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.</p> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.3: </span>The import system has been updated to fully implement the second phase of <span class="target" id="index-1"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0302"><strong>PEP 302</strong></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"><span class="pre">sys.meta_path</span></code></a>. In addition, native namespace package support has been implemented (see <span class="target" id="index-2"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a>).</p> </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"><span class="pre">importlib</span></code></a><a class="headerlink" href="#importlib" title="Permalink to this headline">¶</a></h2> <p>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"><span class="pre">importlib</span></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"><span class="pre">importlib.import_module()</span></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"><span class="pre">__import__()</span></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"><span class="pre">importlib</span></code></a> library documentation for additional detail.</p> </div> <div class="section" id="packages"> <h2>5.2. Packages<a class="headerlink" href="#packages" title="Permalink to this headline">¶</a></h2> <p id="index-3">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"><span class="xref std std-term">packages</span></a>.</p> <p>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.</p> <p>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"><span class="pre">__path__</span></code> attribute is considered a package.</p> <p>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"><span class="pre">sys</span></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"><span class="pre">email</span></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"><span class="pre">email.mime</span></code></a> and a module within that subpackage called <code class="xref py py-mod docutils literal notranslate"><span class="pre">email.mime.text</span></code>.</p> <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> <p id="index-4">Python defines two types of packages, <a class="reference internal" href="../glossary.html#term-regular-package"><span class="xref std std-term">regular packages</span></a> and <a class="reference internal" href="../glossary.html#term-namespace-package"><span class="xref std std-term">namespace packages</span></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"><span class="pre">__init__.py</span></code> file. When a regular package is imported, this <code class="docutils literal notranslate"><span class="pre">__init__.py</span></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"><span class="pre">__init__.py</span></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.</p> <p>For example, the following file system layout defines a top level <code class="docutils literal notranslate"><span class="pre">parent</span></code> package with three subpackages:</p> <div class="highlight-python3 notranslate"><div class="highlight"><pre><span></span><span class="n">parent</span><span class="o">/</span> <span class="fm">__init__</span><span class="o">.</span><span class="n">py</span> <span class="n">one</span><span class="o">/</span> <span class="fm">__init__</span><span class="o">.</span><span class="n">py</span> <span class="n">two</span><span class="o">/</span> <span class="fm">__init__</span><span class="o">.</span><span class="n">py</span> <span class="n">three</span><span class="o">/</span> <span class="fm">__init__</span><span class="o">.</span><span class="n">py</span> </pre></div> </div> <p>Importing <code class="docutils literal notranslate"><span class="pre">parent.one</span></code> will implicitly execute <code class="docutils literal notranslate"><span class="pre">parent/__init__.py</span></code> and <code class="docutils literal notranslate"><span class="pre">parent/one/__init__.py</span></code>. Subsequent imports of <code class="docutils literal notranslate"><span class="pre">parent.two</span></code> or <code class="docutils literal notranslate"><span class="pre">parent.three</span></code> will execute <code class="docutils literal notranslate"><span class="pre">parent/two/__init__.py</span></code> and <code class="docutils literal notranslate"><span class="pre">parent/three/__init__.py</span></code> respectively.</p> </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> <p id="index-5">A namespace package is a composite of various <a class="reference internal" href="../glossary.html#term-portion"><span class="xref std std-term">portions</span></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.</p> <p>Namespace packages do not use an ordinary list for their <code class="docutils literal notranslate"><span class="pre">__path__</span></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"><span class="pre">sys.path</span></code></a> for a top level package) changes.</p> <p>With namespace packages, there is no <code class="docutils literal notranslate"><span class="pre">parent/__init__.py</span></code> file. In fact, there may be multiple <code class="docutils literal notranslate"><span class="pre">parent</span></code> directories found during import search, where each one is provided by a different portion. Thus <code class="docutils literal notranslate"><span class="pre">parent/one</span></code> may not be physically located next to <code class="docutils literal notranslate"><span class="pre">parent/two</span></code>. In this case, Python will create a namespace package for the top-level <code class="docutils literal notranslate"><span class="pre">parent</span></code> package whenever it or one of its subpackages is imported.</p> <p>See also <span class="target" id="index-6"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a> for the namespace package specification.</p> </div> </div> <div class="section" id="searching"> <h2>5.3. Searching<a class="headerlink" href="#searching" title="Permalink to this headline">¶</a></h2> <p>To begin the search, Python needs the <a class="reference internal" href="../glossary.html#term-qualified-name"><span class="xref std std-term">fully qualified</span></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"><span class="pre">import</span></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"><span class="pre">importlib.import_module()</span></code></a> or <a class="reference internal" href="../library/functions.html#__import__" title="__import__"><code class="xref py py-func docutils literal notranslate"><span class="pre">__import__()</span></code></a> functions.</p> <p>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"><span class="pre">foo.bar.baz</span></code>. In this case, Python first tries to import <code class="docutils literal notranslate"><span class="pre">foo</span></code>, then <code class="docutils literal notranslate"><span class="pre">foo.bar</span></code>, and finally <code class="docutils literal notranslate"><span class="pre">foo.bar.baz</span></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"><span class="pre">ModuleNotFoundError</span></code></a> is raised.</p> <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> <p id="index-7">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"><span class="pre">sys.modules</span></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"><span class="pre">foo.bar.baz</span></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"><span class="pre">sys.modules</span></code></a> will contain entries for <code class="docutils literal notranslate"><span class="pre">foo</span></code>, <code class="docutils literal notranslate"><span class="pre">foo.bar</span></code>, and <code class="docutils literal notranslate"><span class="pre">foo.bar.baz</span></code>. Each key will have as its value the corresponding module object.</p> <p>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"><span class="pre">sys.modules</span></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"><span class="pre">None</span></code>, then a <a class="reference internal" href="../library/exceptions.html#ModuleNotFoundError" title="ModuleNotFoundError"><code class="xref py py-exc docutils literal notranslate"><span class="pre">ModuleNotFoundError</span></code></a> is raised. If the module name is missing, Python will continue searching for the module.</p> <p><a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate"><span class="pre">sys.modules</span></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"><span class="pre">None</span></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"><span class="pre">ModuleNotFoundError</span></code></a>.</p> <p>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"><span class="pre">sys.modules</span></code></a>, and then re-import the named module, the two module objects will <em>not</em> 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"><span class="pre">importlib.reload()</span></code></a> will reuse the <em>same</em> module object, and simply reinitialise the module contents by rerunning the module’s code.</p> </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> <p id="index-8">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"><span class="pre">sys.modules</span></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"><span class="xref std std-term">finders</span></a> and <a class="reference internal" href="../glossary.html#term-loader"><span class="xref std std-term">loaders</span></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"><span class="xref std std-term">importers</span></a> - they return themselves when they find that they can load the requested module.</p> <p>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"><span class="xref std std-term">import path</span></a> for modules. The <a class="reference internal" href="../glossary.html#term-import-path"><span class="xref std std-term">import path</span></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.</p> <p>The import machinery is extensible, so new finders can be added to extend the range and scope of module searching.</p> <p>Finders do not actually load modules. If they can find the named module, they return a <em class="dfn">module spec</em>, an encapsulation of the module’s import-related information, which the import machinery then uses when loading the module.</p> <p>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.</p> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.4: </span>In previous versions of Python, finders returned <a class="reference internal" href="../glossary.html#term-loader"><span class="xref std std-term">loaders</span></a> directly, whereas now they return module specs which <em>contain</em> loaders. Loaders are still used during import but have fewer responsibilities.</p> </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> <p id="index-9">The import machinery is designed to be extensible; the primary mechanism for this are the <em>import hooks</em>. There are two types of import hooks: <em>meta hooks</em> and <em>import path hooks</em>.</p> <p>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"><span class="pre">sys.modules</span></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"><span class="pre">sys.path</span></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"><span class="pre">sys.meta_path</span></code></a>, as described below.</p> <p>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"><span class="pre">sys.path</span></code></a> (or <code class="docutils literal notranslate"><span class="pre">package.__path__</span></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"><span class="pre">sys.path_hooks</span></code></a> as described below.</p> </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> <p id="index-10">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"><span class="pre">sys.modules</span></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"><span class="pre">sys.meta_path</span></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"><span class="pre">find_spec()</span></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.</p> <p>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"><span class="pre">None</span></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"><span class="pre">sys.meta_path</span></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"><span class="pre">ModuleNotFoundError</span></code></a> is raised. Any other exceptions raised are simply propagated up, aborting the import process.</p> <p>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"><span class="pre">find_spec()</span></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"><span class="pre">foo.bar.baz</span></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"><span class="pre">None</span></code>, but for submodules or subpackages, the second argument is the value of the parent package’s <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attribute. If the appropriate <code class="docutils literal notranslate"><span class="pre">__path__</span></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"><span class="pre">ModuleNotFoundError</span></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.</p> <p>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"><span class="pre">foo.bar.baz</span></code> will first perform a top level import, calling <code class="docutils literal notranslate"><span class="pre">mpf.find_spec("foo",</span> <span class="pre">None,</span> <span class="pre">None)</span></code> on each meta path finder (<code class="docutils literal notranslate"><span class="pre">mpf</span></code>). After <code class="docutils literal notranslate"><span class="pre">foo</span></code> has been imported, <code class="docutils literal notranslate"><span class="pre">foo.bar</span></code> will be imported by traversing the meta path a second time, calling <code class="docutils literal notranslate"><span class="pre">mpf.find_spec("foo.bar",</span> <span class="pre">foo.__path__,</span> <span class="pre">None)</span></code>. Once <code class="docutils literal notranslate"><span class="pre">foo.bar</span></code> has been imported, the final traversal will call <code class="docutils literal notranslate"><span class="pre">mpf.find_spec("foo.bar.baz",</span> <span class="pre">foo.bar.__path__,</span> <span class="pre">None)</span></code>.</p> <p>Some meta path finders only support top level imports. These importers will always return <code class="docutils literal notranslate"><span class="pre">None</span></code> when anything other than <code class="docutils literal notranslate"><span class="pre">None</span></code> is passed as the second argument.</p> <p>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"><span class="pre">sys.meta_path</span></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"><span class="xref std std-term">import path</span></a> (i.e. the <a class="reference internal" href="../glossary.html#term-path-based-finder"><span class="xref std std-term">path based finder</span></a>).</p> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.4: </span>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"><span class="pre">find_spec()</span></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"><span class="pre">find_module()</span></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"><span class="pre">find_spec()</span></code>.</p> </div> </div> </div> <div class="section" id="loading"> <h2>5.4. Loading<a class="headerlink" href="#loading" title="Permalink to this headline">¶</a></h2> <p>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:</p> <div class="highlight-python3 notranslate"><div class="highlight"><pre><span></span><span class="n">module</span> <span class="o">=</span> <span class="kc">None</span> <span class="k">if</span> <span class="n">spec</span><span class="o">.</span><span class="n">loader</span> <span class="ow">is</span> <span class="ow">not</span> <span class="kc">None</span> <span class="ow">and</span> <span class="nb">hasattr</span><span class="p">(</span><span class="n">spec</span><span class="o">.</span><span class="n">loader</span><span class="p">,</span> <span class="s1">'create_module'</span><span class="p">):</span> <span class="c1"># It is assumed 'exec_module' will also be defined on the loader.</span> <span class="n">module</span> <span class="o">=</span> <span class="n">spec</span><span class="o">.</span><span class="n">loader</span><span class="o">.</span><span class="n">create_module</span><span class="p">(</span><span class="n">spec</span><span class="p">)</span> <span class="k">if</span> <span class="n">module</span> <span class="ow">is</span> <span class="kc">None</span><span class="p">:</span> <span class="n">module</span> <span class="o">=</span> <span class="n">ModuleType</span><span class="p">(</span><span class="n">spec</span><span class="o">.</span><span class="n">name</span><span class="p">)</span> <span class="c1"># The import-related module attributes get set here:</span> <span class="n">_init_module_attrs</span><span class="p">(</span><span class="n">spec</span><span class="p">,</span> <span class="n">module</span><span class="p">)</span> <span class="k">if</span> <span class="n">spec</span><span class="o">.</span><span class="n">loader</span> <span class="ow">is</span> <span class="kc">None</span><span class="p">:</span> <span class="k">if</span> <span class="n">spec</span><span class="o">.</span><span class="n">submodule_search_locations</span> <span class="ow">is</span> <span class="ow">not</span> <span class="kc">None</span><span class="p">:</span> <span class="c1"># namespace package</span> <span class="n">sys</span><span class="o">.</span><span class="n">modules</span><span class="p">[</span><span class="n">spec</span><span class="o">.</span><span class="n">name</span><span class="p">]</span> <span class="o">=</span> <span class="n">module</span> <span class="k">else</span><span class="p">:</span> <span class="c1"># unsupported</span> <span class="k">raise</span> <span class="ne">ImportError</span> <span class="k">elif</span> <span class="ow">not</span> <span class="nb">hasattr</span><span class="p">(</span><span class="n">spec</span><span class="o">.</span><span class="n">loader</span><span class="p">,</span> <span class="s1">'exec_module'</span><span class="p">):</span> <span class="n">module</span> <span class="o">=</span> <span class="n">spec</span><span class="o">.</span><span class="n">loader</span><span class="o">.</span><span class="n">load_module</span><span class="p">(</span><span class="n">spec</span><span class="o">.</span><span class="n">name</span><span class="p">)</span> <span class="c1"># Set __loader__ and __package__ if missing.</span> <span class="k">else</span><span class="p">:</span> <span class="n">sys</span><span class="o">.</span><span class="n">modules</span><span class="p">[</span><span class="n">spec</span><span class="o">.</span><span class="n">name</span><span class="p">]</span> <span class="o">=</span> <span class="n">module</span> <span class="k">try</span><span class="p">:</span> <span class="n">spec</span><span class="o">.</span><span class="n">loader</span><span class="o">.</span><span class="n">exec_module</span><span class="p">(</span><span class="n">module</span><span class="p">)</span> <span class="k">except</span> <span class="ne">BaseException</span><span class="p">:</span> <span class="k">try</span><span class="p">:</span> <span class="k">del</span> <span class="n">sys</span><span class="o">.</span><span class="n">modules</span><span class="p">[</span><span class="n">spec</span><span class="o">.</span><span class="n">name</span><span class="p">]</span> <span class="k">except</span> <span class="ne">KeyError</span><span class="p">:</span> <span class="k">pass</span> <span class="k">raise</span> <span class="k">return</span> <span class="n">sys</span><span class="o">.</span><span class="n">modules</span><span class="p">[</span><span class="n">spec</span><span class="o">.</span><span class="n">name</span><span class="p">]</span> </pre></div> </div> <p>Note the following details:</p> <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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></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"><span class="std std-ref">later section</span></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"> <p><span class="versionmodified">Changed in version 3.4: </span>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"><span class="pre">importlib.abc.Loader.load_module()</span></code></a> method.</p> </div> <div class="section" id="loaders"> <h3>5.4.1. Loaders<a class="headerlink" href="#loaders" title="Permalink to this headline">¶</a></h3> <p>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"><span class="pre">importlib.abc.Loader.exec_module()</span></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"><span class="pre">exec_module()</span></code></a> is ignored.</p> <p>Loaders must satisfy the following requirements:</p> <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"><span class="pre">module.__dict__</span></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"><span class="pre">ImportError</span></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"><span class="pre">exec_module()</span></code></a> will be propagated.</li> </ul> </div></blockquote> <p>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"><span class="pre">find_spec()</span></code></a> method would just return a spec with the loader set to <code class="docutils literal notranslate"><span class="pre">self</span></code>.</p> <p>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"><span class="pre">create_module()</span></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"><span class="pre">create_module()</span></code> does not need to set any attributes on the module object. If the method returns <code class="docutils literal notranslate"><span class="pre">None</span></code>, the import machinery will create the new module itself.</p> <div class="versionadded"> <p><span class="versionmodified">New in version 3.4: </span>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"><span class="pre">create_module()</span></code></a> method of loaders.</p> </div> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.4: </span>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"><span class="pre">load_module()</span></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"><span class="pre">exec_module()</span></code></a> and the import machinery assumed all the boilerplate responsibilities of loading.</p> <p>For compatibility with existing loaders, the import machinery will use the <code class="docutils literal notranslate"><span class="pre">load_module()</span></code> method of loaders if it exists and the loader does not also implement <code class="docutils literal notranslate"><span class="pre">exec_module()</span></code>. However, <code class="docutils literal notranslate"><span class="pre">load_module()</span></code> has been deprecated and loaders should implement <code class="docutils literal notranslate"><span class="pre">exec_module()</span></code> instead.</p> <p>The <code class="docutils literal notranslate"><span class="pre">load_module()</span></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:</p> <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"><span class="pre">sys.modules</span></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"><span class="pre">importlib.reload()</span></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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></code></a>.</li> <li>The module <em>must</em> exist in <a class="reference internal" href="../library/sys.html#sys.modules" title="sys.modules"><code class="xref py py-data docutils literal notranslate"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></code></a>, but it must remove <strong>only</strong> the failing module(s), and only if the loader itself has loaded the module(s) explicitly.</li> </ul> </div></blockquote> </div> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.5: </span>A <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate"><span class="pre">DeprecationWarning</span></code></a> is raised when <code class="docutils literal notranslate"><span class="pre">exec_module()</span></code> is defined but <code class="docutils literal notranslate"><span class="pre">create_module()</span></code> is not.</p> </div> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.6: </span>An <a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate"><span class="pre">ImportError</span></code></a> is raised when <code class="docutils literal notranslate"><span class="pre">exec_module()</span></code> is defined but <code class="docutils literal notranslate"><span class="pre">create_module()</span></code> is not.</p> </div> </div> <div class="section" id="submodules"> <h3>5.4.2. Submodules<a class="headerlink" href="#submodules" title="Permalink to this headline">¶</a></h3> <p>When a submodule is loaded using any mechanism (e.g. <code class="docutils literal notranslate"><span class="pre">importlib</span></code> APIs, the <code class="docutils literal notranslate"><span class="pre">import</span></code> or <code class="docutils literal notranslate"><span class="pre">import-from</span></code> statements, or built-in <code class="docutils literal notranslate"><span class="pre">__import__()</span></code>) a binding is placed in the parent module’s namespace to the submodule object. For example, if package <code class="docutils literal notranslate"><span class="pre">spam</span></code> has a submodule <code class="docutils literal notranslate"><span class="pre">foo</span></code>, after importing <code class="docutils literal notranslate"><span class="pre">spam.foo</span></code>, <code class="docutils literal notranslate"><span class="pre">spam</span></code> will have an attribute <code class="docutils literal notranslate"><span class="pre">foo</span></code> which is bound to the submodule. Let’s say you have the following directory structure:</p> <div class="highlight-python3 notranslate"><div class="highlight"><pre><span></span><span class="n">spam</span><span class="o">/</span> <span class="fm">__init__</span><span class="o">.</span><span class="n">py</span> <span class="n">foo</span><span class="o">.</span><span class="n">py</span> <span class="n">bar</span><span class="o">.</span><span class="n">py</span> </pre></div> </div> <p>and <code class="docutils literal notranslate"><span class="pre">spam/__init__.py</span></code> has the following lines in it:</p> <div class="highlight-python3 notranslate"><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">.foo</span> <span class="k">import</span> <span class="n">Foo</span> <span class="kn">from</span> <span class="nn">.bar</span> <span class="k">import</span> <span class="n">Bar</span> </pre></div> </div> <p>then executing the following puts a name binding to <code class="docutils literal notranslate"><span class="pre">foo</span></code> and <code class="docutils literal notranslate"><span class="pre">bar</span></code> in the <code class="docutils literal notranslate"><span class="pre">spam</span></code> module:</p> <div class="highlight-python3 notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="kn">import</span> <span class="nn">spam</span> <span class="gp">>>> </span><span class="n">spam</span><span class="o">.</span><span class="n">foo</span> <span class="go"><module 'spam.foo' from '/tmp/imports/spam/foo.py'></span> <span class="gp">>>> </span><span class="n">spam</span><span class="o">.</span><span class="n">bar</span> <span class="go"><module 'spam.bar' from '/tmp/imports/spam/bar.py'></span> </pre></div> </div> <p>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"><span class="pre">sys.modules['spam']</span></code> and <code class="docutils literal notranslate"><span class="pre">sys.modules['spam.foo']</span></code> (as you would after the above import), the latter must appear as the <code class="docutils literal notranslate"><span class="pre">foo</span></code> attribute of the former.</p> </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> <p>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.</p> <p>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.</p> <p>The module’s spec is exposed as the <code class="docutils literal notranslate"><span class="pre">__spec__</span></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"><span class="pre">ModuleSpec</span></code></a> for details on the contents of the module spec.</p> <div class="versionadded"> <p><span class="versionmodified">New in version 3.4.</span></p> </div> </div> <div class="section" id="import-related-module-attributes"> <span id="import-mod-attrs"></span><h3>5.4.4. Import-related module attributes<a class="headerlink" href="#import-related-module-attributes" title="Permalink to this headline">¶</a></h3> <p>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.</p> <dl class="attribute"> <dt id="__name__"> <code class="descname">__name__</code><a class="headerlink" href="#__name__" title="Permalink to this definition">¶</a></dt> <dd><p>The <code class="docutils literal notranslate"><span class="pre">__name__</span></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.</p> </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><p>The <code class="docutils literal notranslate"><span class="pre">__loader__</span></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.</p> </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><p>The module’s <code class="docutils literal notranslate"><span class="pre">__package__</span></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"><span class="pre">__name__</span></code>. When the module is a package, its <code class="docutils literal notranslate"><span class="pre">__package__</span></code> value should be set to its <code class="docutils literal notranslate"><span class="pre">__name__</span></code>. When the module is not a package, <code class="docutils literal notranslate"><span class="pre">__package__</span></code> should be set to the empty string for top-level modules, or for submodules, to the parent package’s name. See <span class="target" id="index-11"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366"><strong>PEP 366</strong></a> for further details.</p> <p>This attribute is used instead of <code class="docutils literal notranslate"><span class="pre">__name__</span></code> to calculate explicit relative imports for main modules, as defined in <span class="target" id="index-12"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366"><strong>PEP 366</strong></a>. It is expected to have the same value as <code class="docutils literal notranslate"><span class="pre">__spec__.parent</span></code>.</p> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.6: </span>The value of <code class="docutils literal notranslate"><span class="pre">__package__</span></code> is expected to be the same as <code class="docutils literal notranslate"><span class="pre">__spec__.parent</span></code>.</p> </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><p>The <code class="docutils literal notranslate"><span class="pre">__spec__</span></code> attribute must be set to the module spec that was used when importing the module. Setting <code class="docutils literal notranslate"><span class="pre">__spec__</span></code> appropriately applies equally to <a class="reference internal" href="toplevel_components.html#programs"><span class="std std-ref">modules initialized during interpreter startup</span></a>. The one exception is <code class="docutils literal notranslate"><span class="pre">__main__</span></code>, where <code class="docutils literal notranslate"><span class="pre">__spec__</span></code> is <a class="reference internal" href="#main-spec"><span class="std std-ref">set to None in some cases</span></a>.</p> <p>When <code class="docutils literal notranslate"><span class="pre">__package__</span></code> is not defined, <code class="docutils literal notranslate"><span class="pre">__spec__.parent</span></code> is used as a fallback.</p> <div class="versionadded"> <p><span class="versionmodified">New in version 3.4.</span></p> </div> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.6: </span><code class="docutils literal notranslate"><span class="pre">__spec__.parent</span></code> is used as a fallback when <code class="docutils literal notranslate"><span class="pre">__package__</span></code> is not defined.</p> </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><p>If the module is a package (either regular or namespace), the module object’s <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attribute must be set. The value must be iterable, but may be empty if <code class="docutils literal notranslate"><span class="pre">__path__</span></code> has no further significance. If <code class="docutils literal notranslate"><span class="pre">__path__</span></code> is not empty, it must produce strings when iterated over. More details on the semantics of <code class="docutils literal notranslate"><span class="pre">__path__</span></code> are given <a class="reference internal" href="#package-path-rules"><span class="std std-ref">below</span></a>.</p> <p>Non-package modules should not have a <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attribute.</p> </dd></dl> <dl class="attribute"> <dt id="__file__"> <code class="descname">__file__</code><a class="headerlink" href="#__file__" title="Permalink to this definition">¶</a></dt> <dd></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><p><code class="docutils literal notranslate"><span class="pre">__file__</span></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"><span class="pre">__file__</span></code> unset if it has no semantic meaning (e.g. a module loaded from a database).</p> <p>If <code class="docutils literal notranslate"><span class="pre">__file__</span></code> is set, it may also be appropriate to set the <code class="docutils literal notranslate"><span class="pre">__cached__</span></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 <span class="target" id="index-13"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-3147"><strong>PEP 3147</strong></a>).</p> <p>It is also appropriate to set <code class="docutils literal notranslate"><span class="pre">__cached__</span></code> when <code class="docutils literal notranslate"><span class="pre">__file__</span></code> is not set. However, that scenario is quite atypical. Ultimately, the loader is what makes use of <code class="docutils literal notranslate"><span class="pre">__file__</span></code> and/or <code class="docutils literal notranslate"><span class="pre">__cached__</span></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.</p> </dd></dl> </div> <div class="section" id="module-path"> <span id="package-path-rules"></span><h3>5.4.5. module.__path__<a class="headerlink" href="#module-path" title="Permalink to this headline">¶</a></h3> <p>By definition, if a module has a <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attribute, it is a package.</p> <p>A package’s <code class="docutils literal notranslate"><span class="pre">__path__</span></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"><span class="pre">sys.path</span></code></a>, i.e. providing a list of locations to search for modules during import. However, <code class="docutils literal notranslate"><span class="pre">__path__</span></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"><span class="pre">sys.path</span></code></a>.</p> <p><code class="docutils literal notranslate"><span class="pre">__path__</span></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"><span class="pre">sys.path</span></code></a> also apply to a package’s <code class="docutils literal notranslate"><span class="pre">__path__</span></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"><span class="pre">sys.path_hooks</span></code></a> (described below) are consulted when traversing a package’s <code class="docutils literal notranslate"><span class="pre">__path__</span></code>.</p> <p>A package’s <code class="docutils literal notranslate"><span class="pre">__init__.py</span></code> file may set or alter the package’s <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attribute, and this was typically the way namespace packages were implemented prior to <span class="target" id="index-14"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a>. With the adoption of <span class="target" id="index-15"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a>, namespace packages no longer need to supply <code class="docutils literal notranslate"><span class="pre">__init__.py</span></code> files containing only <code class="docutils literal notranslate"><span class="pre">__path__</span></code> manipulation code; the import machinery automatically sets <code class="docutils literal notranslate"><span class="pre">__path__</span></code> correctly for the namespace package.</p> </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> <p>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.</p> <p>If the module has a spec (<code class="docutils literal notranslate"><span class="pre">__spec__</span></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"><span class="pre">module.__name__</span></code>, <code class="docutils literal notranslate"><span class="pre">module.__file__</span></code>, and <code class="docutils literal notranslate"><span class="pre">module.__loader__</span></code> as input into the repr, with defaults for whatever information is missing.</p> <p>Here are the exact rules used:</p> <blockquote> <div><ul class="simple"> <li>If the module has a <code class="docutils literal notranslate"><span class="pre">__spec__</span></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"><span class="pre">__file__</span></code> attribute, this is used as part of the module’s repr.</li> <li>If the module has no <code class="docutils literal notranslate"><span class="pre">__file__</span></code> but does have a <code class="docutils literal notranslate"><span class="pre">__loader__</span></code> that is not <code class="docutils literal notranslate"><span class="pre">None</span></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"><span class="pre">__name__</span></code> in the repr.</li> </ul> </div></blockquote> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.4: </span>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"><span class="pre">loader.module_repr()</span></code></a> has been deprecated and the module spec is now used by the import machinery to generate a module repr.</p> <p>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"><span class="pre">module_repr()</span></code></a> method, if defined, before trying either approach described above. However, the method is deprecated.</p> </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> <p id="index-16">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"><span class="xref std std-term">path based finder</span></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"><span class="pre">PathFinder</span></code></a>), searches an <a class="reference internal" href="../glossary.html#term-import-path"><span class="xref std std-term">import path</span></a>, which contains a list of <a class="reference internal" href="../glossary.html#term-path-entry"><span class="xref std std-term">path entries</span></a>. Each path entry names a location to search for modules.</p> <p>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.</p> <p>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"><span class="pre">.py</span></code> files), Python byte code (<code class="docutils literal notranslate"><span class="pre">.pyc</span></code> files) and shared libraries (e.g. <code class="docutils literal notranslate"><span class="pre">.so</span></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"><span class="pre">zipimport</span></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.</p> <p>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.</p> <p>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"><span class="xref std std-term">path entry finder</span></a> supporting the protocol described below, which was then used to get a loader for the module from the web.</p> <p>A word of warning: this section and the previous both use the term <em>finder</em>, distinguishing between them by using the terms <a class="reference internal" href="../glossary.html#term-meta-path-finder"><span class="xref std std-term">meta path finder</span></a> and <a class="reference internal" href="../glossary.html#term-path-entry-finder"><span class="xref std std-term">path entry finder</span></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"><span class="pre">sys.meta_path</span></code></a> traversal.</p> <p>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"><span class="pre">sys.meta_path</span></code></a>, none of the path entry finder semantics would be invoked.</p> <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> <p id="index-17">The <a class="reference internal" href="../glossary.html#term-path-based-finder"><span class="xref std std-term">path based finder</span></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"><span class="xref std std-term">path entry</span></a>. Most path entries name locations in the file system, but they need not be limited to this.</p> <p>As a meta path finder, the <a class="reference internal" href="../glossary.html#term-path-based-finder"><span class="xref std std-term">path based finder</span></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"><span class="pre">find_spec()</span></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"><span class="xref std std-term">import path</span></a>.</p> <p>Three variables are used by the <a class="reference internal" href="../glossary.html#term-path-based-finder"><span class="xref std std-term">path based finder</span></a>, <a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate"><span class="pre">sys.path</span></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"><span class="pre">sys.path_hooks</span></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"><span class="pre">sys.path_importer_cache</span></code></a>. The <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attributes on package objects are also used. These provide additional ways that the import machinery can be customized.</p> <p><a class="reference internal" href="../library/sys.html#sys.path" title="sys.path"><code class="xref py py-data docutils literal notranslate"><span class="pre">sys.path</span></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"><span class="pre">PYTHONPATH</span></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"><span class="pre">sys.path</span></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"><span class="pre">site</span></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"><span class="pre">sys.path</span></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"><span class="xref std std-term">path entry finders</span></a>.</p> <p>The <a class="reference internal" href="../glossary.html#term-path-based-finder"><span class="xref std std-term">path based finder</span></a> is a <a class="reference internal" href="../glossary.html#term-meta-path-finder"><span class="xref std std-term">meta path finder</span></a>, so the import machinery begins the <a class="reference internal" href="../glossary.html#term-import-path"><span class="xref std std-term">import path</span></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"><span class="pre">find_spec()</span></code></a> method as described previously. When the <code class="docutils literal notranslate"><span class="pre">path</span></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"><span class="pre">find_spec()</span></code></a> is given, it will be a list of string paths to traverse - typically a package’s <code class="docutils literal notranslate"><span class="pre">__path__</span></code> attribute for an import within that package. If the <code class="docutils literal notranslate"><span class="pre">path</span></code> argument is <code class="docutils literal notranslate"><span class="pre">None</span></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"><span class="pre">sys.path</span></code></a> is used.</p> <p>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"><span class="xref std std-term">path entry finder</span></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"><span class="pre">PathEntryFinder</span></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"><span class="pre">sys.path_importer_cache</span></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"><span class="xref std std-term">importer</span></a> objects). In this way, the expensive search for a particular <a class="reference internal" href="../glossary.html#term-path-entry"><span class="xref std std-term">path entry</span></a> location’s <a class="reference internal" href="../glossary.html#term-path-entry-finder"><span class="xref std std-term">path entry finder</span></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"><span class="pre">sys.path_importer_cache</span></code></a> forcing the path based finder to perform the path entry search again <a class="footnote-reference" href="#fnpic" id="id3">[3]</a>.</p> <p>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"><span class="pre">sys.path_hooks</span></code></a>. Each of the <a class="reference internal" href="../glossary.html#term-path-entry-hook"><span class="xref std std-term">path entry hooks</span></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"><span class="xref std std-term">path entry finder</span></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"><span class="pre">ImportError</span></code></a>. An <a class="reference internal" href="../library/exceptions.html#ImportError" title="ImportError"><code class="xref py py-exc docutils literal notranslate"><span class="pre">ImportError</span></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"><span class="xref std std-term">path entry finder</span></a> for that <a class="reference internal" href="../glossary.html#term-path-entry"><span class="xref std std-term">path entry</span></a>. The exception is ignored and <a class="reference internal" href="../glossary.html#term-import-path"><span class="xref std std-term">import path</span></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"><span class="pre">ImportError</span></code></a>.</p> <p>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"><span class="pre">sys.path_hooks</span></code></a> iteration ends with no <a class="reference internal" href="../glossary.html#term-path-entry-finder"><span class="xref std std-term">path entry finder</span></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"><span class="pre">find_spec()</span></code></a> method will store <code class="docutils literal notranslate"><span class="pre">None</span></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"><span class="pre">sys.path_importer_cache</span></code></a> (to indicate that there is no finder for this path entry) and return <code class="docutils literal notranslate"><span class="pre">None</span></code>, indicating that this <a class="reference internal" href="../glossary.html#term-meta-path-finder"><span class="xref std std-term">meta path finder</span></a> could not find the module.</p> <p>If a <a class="reference internal" href="../glossary.html#term-path-entry-finder"><span class="xref std std-term">path entry finder</span></a> <em>is</em> returned by one of the <a class="reference internal" href="../glossary.html#term-path-entry-hook"><span class="xref std std-term">path entry hook</span></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"><span class="pre">sys.path_hooks</span></code></a>, then the following protocol is used to ask the finder for a module spec, which is then used when loading the module.</p> <p>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"><span class="pre">sys.path</span></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"><span class="pre">sys.path_importer_cache</span></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"><span class="pre">sys.path_importer_cache</span></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"><span class="pre">importlib.machinery.PathFinder.find_spec()</span></code></a> will be the actual current working directory and not the empty string.</p> </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> <p>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"><span class="pre">find_spec()</span></code></a> method.</p> <p><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"><span class="pre">find_spec()</span></code></a> takes two argument, the fully qualified name of the module being imported, and the (optional) target module. <code class="docutils literal notranslate"><span class="pre">find_spec()</span></code> returns a fully populated spec for the module. This spec will always have “loader” set (with one exception).</p> <p>To indicate to the import machinery that the spec represents a namespace <a class="reference internal" href="../glossary.html#term-portion"><span class="xref std std-term">portion</span></a>. the path entry finder sets “loader” on the spec to <code class="docutils literal notranslate"><span class="pre">None</span></code> and “submodule_search_locations” to a list containing the portion.</p> <div class="versionchanged"> <p><span class="versionmodified">Changed in version 3.4: </span><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"><span class="pre">find_spec()</span></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"><span class="pre">find_loader()</span></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"><span class="pre">find_module()</span></code></a>, both of which are now deprecated, but will be used if <code class="docutils literal notranslate"><span class="pre">find_spec()</span></code> is not defined.</p> <p>Older path entry finders may implement one of these two deprecated methods instead of <code class="docutils literal notranslate"><span class="pre">find_spec()</span></code>. The methods are still respected for the sake of backward compatibility. However, if <code class="docutils literal notranslate"><span class="pre">find_spec()</span></code> is implemented on the path entry finder, the legacy methods are ignored.</p> <p><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"><span class="pre">find_loader()</span></code></a> takes one argument, the fully qualified name of the module being imported. <code class="docutils literal notranslate"><span class="pre">find_loader()</span></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"><span class="xref std std-term">portion</span></a>. When the first item (i.e. the loader) is <code class="docutils literal notranslate"><span class="pre">None</span></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"><span class="pre">None</span></code> for the loader, the second item of the 2-tuple return value must be a sequence, although it can be empty.</p> <p>If <code class="docutils literal notranslate"><span class="pre">find_loader()</span></code> returns a non-<code class="docutils literal notranslate"><span class="pre">None</span></code> loader value, the portion is ignored and the loader is returned from the path based finder, terminating the search through the path entries.</p> <p>For backwards compatibility with other implementations of the import protocol, many path entry finders also support the same, traditional <code class="docutils literal notranslate"><span class="pre">find_module()</span></code> method that meta path finders support. However path entry finder <code class="docutils literal notranslate"><span class="pre">find_module()</span></code> methods are never called with a <code class="docutils literal notranslate"><span class="pre">path</span></code> argument (they are expected to record the appropriate path information from the initial call to the path hook).</p> <p>The <code class="docutils literal notranslate"><span class="pre">find_module()</span></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"><span class="pre">find_loader()</span></code> and <code class="docutils literal notranslate"><span class="pre">find_module()</span></code> exist on a path entry finder, the import system will always call <code class="docutils literal notranslate"><span class="pre">find_loader()</span></code> in preference to <code class="docutils literal notranslate"><span class="pre">find_module()</span></code>.</p> </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> <p>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"><span class="pre">sys.meta_path</span></code></a>, replacing them entirely with a custom meta path hook.</p> <p>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"><span class="pre">__import__()</span></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.</p> <p>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"><span class="pre">ModuleNotFoundError</span></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"><span class="pre">find_spec()</span></code></a> instead of returning <code class="docutils literal notranslate"><span class="pre">None</span></code>. The latter indicates that the meta path search should continue, while raising an exception terminates it immediately.</p> </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> <p>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"><span class="pre">__main__</span></code></a> module is a special case relative to Python’s import system. As noted <a class="reference internal" href="toplevel_components.html#programs"><span class="std std-ref">elsewhere</span></a>, the <code class="docutils literal notranslate"><span class="pre">__main__</span></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"><span class="pre">sys</span></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"><span class="pre">builtins</span></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"><span class="pre">__main__</span></code> is initialized depends on the flags and other options with which the interpreter is invoked.</p> <div class="section" id="main-spec"> <span id="id4"></span><h3>5.7.1. __main__.__spec__<a class="headerlink" href="#main-spec" title="Permalink to this headline">¶</a></h3> <p>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"><span class="pre">__main__</span></code></a> is initialized, <code class="docutils literal notranslate"><span class="pre">__main__.__spec__</span></code> gets set appropriately or to <code class="docutils literal notranslate"><span class="pre">None</span></code>.</p> <p>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"><span class="pre">-m</span></code></a> option, <code class="docutils literal notranslate"><span class="pre">__spec__</span></code> is set to the module spec of the corresponding module or package. <code class="docutils literal notranslate"><span class="pre">__spec__</span></code> is also populated when the <code class="docutils literal notranslate"><span class="pre">__main__</span></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"><span class="pre">sys.path</span></code></a> entry.</p> <p>In <a class="reference internal" href="../using/cmdline.html#using-on-interface-options"><span class="std std-ref">the remaining cases</span></a> <code class="docutils literal notranslate"><span class="pre">__main__.__spec__</span></code> is set to <code class="docutils literal notranslate"><span class="pre">None</span></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"><span class="pre">__main__</span></code></a> does not correspond directly with an importable module:</p> <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> <p>Note that <code class="docutils literal notranslate"><span class="pre">__main__.__spec__</span></code> is always <code class="docutils literal notranslate"><span class="pre">None</span></code> in the last case, <em>even if</em> 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"><span class="pre">-m</span></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"><span class="pre">__main__</span></code></a>.</p> <p>Note also that even when <code class="docutils literal notranslate"><span class="pre">__main__</span></code> corresponds with an importable module and <code class="docutils literal notranslate"><span class="pre">__main__.__spec__</span></code> is set accordingly, they’re still considered <em>distinct</em> modules. This is due to the fact that blocks guarded by <code class="docutils literal notranslate"><span class="pre">if</span> <span class="pre">__name__</span> <span class="pre">==</span> <span class="pre">"__main__":</span></code> checks only execute when the module is used to populate the <code class="docutils literal notranslate"><span class="pre">__main__</span></code> namespace, and not during normal import.</p> </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> <p>XXX It would be really nice to have a diagram.</p> <p>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?</p> <p>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.</p> <p>XXX Add more explanation regarding the different ways in which <code class="docutils literal notranslate"><span class="pre">__main__</span></code> is initialized?</p> <p>XXX Add more info on <code class="docutils literal notranslate"><span class="pre">__main__</span></code> quirks/pitfalls (i.e. copy from <span class="target" id="index-18"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0395"><strong>PEP 395</strong></a>).</p> </div> <div class="section" id="references"> <h2>5.9. References<a class="headerlink" href="#references" title="Permalink to this headline">¶</a></h2> <p>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.</p> <p>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"><span class="pre">sys.meta_path</span></code></a> was <span class="target" id="index-19"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0302"><strong>PEP 302</strong></a>, with subsequent extension in <span class="target" id="index-20"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a>.</p> <p><span class="target" id="index-21"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a> introduced <a class="reference internal" href="../glossary.html#term-namespace-package"><span class="xref std std-term">namespace packages</span></a> for Python 3.3. <span class="target" id="index-22"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0420"><strong>PEP 420</strong></a> also introduced the <code class="xref py py-meth docutils literal notranslate"><span class="pre">find_loader()</span></code> protocol as an alternative to <code class="xref py py-meth docutils literal notranslate"><span class="pre">find_module()</span></code>.</p> <p><span class="target" id="index-23"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366"><strong>PEP 366</strong></a> describes the addition of the <code class="docutils literal notranslate"><span class="pre">__package__</span></code> attribute for explicit relative imports in main modules.</p> <p><span class="target" id="index-24"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0328"><strong>PEP 328</strong></a> introduced absolute and explicit relative imports and initially proposed <code class="docutils literal notranslate"><span class="pre">__name__</span></code> for semantics <span class="target" id="index-25"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0366"><strong>PEP 366</strong></a> would eventually specify for <code class="docutils literal notranslate"><span class="pre">__package__</span></code>.</p> <p><span class="target" id="index-26"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0338"><strong>PEP 338</strong></a> defines executing modules as scripts.</p> <p><span class="target" id="index-27"></span><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0451"><strong>PEP 451</strong></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.</p> <p class="rubric">Footnotes</p> <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"><span class="pre">types.ModuleType</span></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"><span class="pre">sys.modules</span></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"><span class="pre">sys.modules</span></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"><span class="pre">imp.NullImporter</span></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"><span class="pre">sys.path_importer_cache</span></code></a>. It is recommended that code be changed to use <code class="docutils literal notranslate"><span class="pre">None</span></code> instead. See <a class="reference internal" href="../whatsnew/3.3.html#portingpythoncode"><span class="std std-ref">Porting Python code</span></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"><span class="pre">importlib</span></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> <p class="topless"><a href="executionmodel.html" title="previous chapter">4. Execution model</a></p> <h4>Next topic</h4> <p class="topless"><a href="expressions.html" title="next chapter">6. Expressions</a></p> <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> »</li> <li> <a href="../index.html">3.6.7 Documentation</a> » </li> <li class="nav-item nav-item-1"><a href="index.html" >The Python Language Reference</a> »</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="footer"> © <a href="../copyright.html">Copyright</a> 2001-2021, Python Software Foundation. <br /> The Python Software Foundation is a non-profit corporation. <a href="https://www.python.org/psf/donations/">Please donate.</a> <br /> Last updated on Aug 25, 2021. <a href="../bugs.html">Found a bug</a>? <br /> Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.7.6. </div> </body> </html>