imp —- Access the import internals

Source code:Lib/imp.py

3.4 版后已移除: The imp package is pending deprecation in favor of importlib.


This module provides an interface to the mechanisms used to implement theimport statement. It defines the following constants and functions:

  • imp.get_magic()
  • Return the magic string value used to recognize byte-compiled code files(.pyc files). (This value may be different for each Python version.)

3.4 版后已移除: Use importlib.util.MAGIC_NUMBER instead.

  • imp.get_suffixes()
  • Return a list of 3-element tuples, each describing a particular type ofmodule. Each triple has the form (suffix, mode, type), where suffix isa string to be appended to the module name to form the filename to searchfor, mode is the mode string to pass to the built-in open() functionto open the file (this can be 'r' for text files or 'rb' for binaryfiles), and type is the file type, which has one of the valuesPY_SOURCE, PY_COMPILED, or C_EXTENSION, describedbelow.

3.3 版后已移除: Use the constants defined on importlib.machinery instead.

  • imp.findmodule(_name[, path])
  • Try to find the module name. If path is omitted or None, the list ofdirectory names given by sys.path is searched, but first a few specialplaces are searched: the function tries to find a built-in module with thegiven name (C_BUILTIN), then a frozen module (PY_FROZEN),and on some systems some other places are looked in as well (on Windows, itlooks in the registry which may point to a specific file).

Otherwise, path must be a list of directory names; each directory issearched for files with any of the suffixes returned by get_suffixes()above. Invalid names in the list are silently ignored (but all list itemsmust be strings).

If search is successful, the return value is a 3-element tuple (file,pathname, description):

file is an open file object positioned at the beginning, pathname_is the pathname of the file found, and _description is a 3-element tuple ascontained in the list returned by get_suffixes() describing the kind ofmodule found.

If the module is built-in or frozen then file and pathname are both Noneand the description tuple contains empty strings for its suffix and mode;the module type is indicated as given in parentheses above. If the searchis unsuccessful, ImportError is raised. Other exceptions indicateproblems with the arguments or environment.

If the module is a package, file is None, pathname is the packagepath and the last item in the description tuple is PKG_DIRECTORY.

This function does not handle hierarchical module names (names containingdots). In order to find P.M, that is, submodule M of package P, usefind_module() and load_module() to find and load package P, andthen use find_module() with the path argument set to P.path.When P itself has a dotted name, apply this recipe recursively.

3.3 版后已移除: Use importlib.util.find_spec() instead unless Python 3.3compatibility is required, in which case useimportlib.find_loader(). For example usage of the former case,see the 示例 section of the importlibdocumentation.

  • imp.loadmodule(_name, file, pathname, description)
  • Load a module that was previously found by find_module() (or by anotherwise conducted search yielding compatible results). This function doesmore than importing the module: if the module was already imported, it willreload the module! The name argument indicates the fullmodule name (including the package name, if this is a submodule of apackage). The file argument is an open file, and pathname is thecorresponding file name; these can be None and '', respectively, whenthe module is a package or not being loaded from a file. The _description_argument is a tuple, as would be returned by get_suffixes(), describingwhat kind of module must be loaded.

If the load is successful, the return value is the module object; otherwise,an exception (usually ImportError) is raised.

Important: the caller is responsible for closing the file argument, ifit was not None, even when an exception is raised. This is best doneusing a tryfinally statement.

3.3 版后已移除: If previously used in conjunction with imp.find_module() thenconsider using importlib.import_module(), otherwise use the loaderreturned by the replacement you chose for imp.find_module(). If youcalled imp.load_module() and related functions directly with filepath arguments then use a combination ofimportlib.util.spec_from_file_location() andimportlib.util.module_from_spec(). See the 示例section of the importlib documentation for details of the variousapproaches.

  • imp.newmodule(_name)
  • Return a new empty module object called name. This object is not insertedin sys.modules.

3.4 版后已移除: Use importlib.util.module_from_spec() instead.

  • imp.reload(module)
  • Reload a previously imported module. The argument must be a module object, soit must have been successfully imported before. This is useful if you haveedited the module source file using an external editor and want to try out thenew version without leaving the Python interpreter. The return value is themodule object (the same as the module argument).

When reload(module) is executed:

  • Python modules' code is recompiled and the module-level code reexecuted,defining a new set of objects which are bound to names in the module'sdictionary. The init function of extension modules is not called a secondtime.

  • 与Python中的所有的其它对象一样,旧的对象只有在引用计数为0之后才会被回收。

  • 模块命名空间中的名称重新指向任何新的或更改后的对象。

  • 其他旧对象的引用(例如那个模块的外部名称)不会被重新绑定到引用的新对象的,并且如果有需要,必须在出现的每个命名空间中进行更新。

有一些其他注意事项:

当一个模块被重新加载的时候,它的字典(包含了那个模块的全区变量)会被保留。名称的重新定义会覆盖旧的定义,所以通常来说这不是问题。如果一个新模块没有定义在旧版本模块中定义的名称,则将保留旧版本中的定义。这一特性可用于作为那个模块的优点,如果它维护一个全局表或者对象的缓存 —— 使用 try 语句,就可以测试表的存在并且跳过它的初始化,如果有需要的话:

  1. try:
  2. cache
  3. except NameError:
  4. cache = {}

It is legal though generally not very useful to reload built-in or dynamicallyloaded modules, except for sys, main and builtins.In many cases, however, extension modules are not designed to be initializedmore than once, and may fail in arbitrary ways when reloaded.

If a module imports objects from another module using fromimport …, calling reload() for the other module does notredefine the objects imported from it —- one way around this is to re-executethe from statement, another is to use import and qualifiednames (module.name) instead.

如果一个模块创建一个类的实例,重新加载定义那个类的模块不影响这个实例的方法定义———它们继续使用旧类的定义。对于子类,也是一样的。

在 3.3 版更改: Relies on both name and loader being defined on the modulebeing reloaded instead of just name.

3.4 版后已移除: Use importlib.reload() instead.

The following functions are conveniences for handling PEP 3147 byte-compiledfile paths.

3.2 新版功能.

  • imp.cachefrom_source(_path, debug_override=None)
  • Return the PEP 3147 path to the byte-compiled file associated with thesource path. For example, if path is /foo/bar/baz.py the returnvalue would be /foo/bar/pycache/baz.cpython-32.pyc for Python 3.2.The cpython-32 string comes from the current magic tag (seeget_tag(); if sys.implementation.cachetag is not defined thenNotImplementedError will be raised). By passing in True orFalse for _debug_override you can override the system's value fordebug, leading to optimized bytecode.

path need not exist.

在 3.3 版更改: If sys.implementation.cache_tag is None, thenNotImplementedError is raised.

3.4 版后已移除: Use importlib.util.cache_from_source() instead.

在 3.5 版更改: The debug_override parameter no longer creates a .pyo file.

  • imp.sourcefrom_cache(_path)
  • Given the path to a PEP 3147 file name, return the associated source codefile path. For example, if path is/foo/bar/pycache/baz.cpython-32.pyc the returned path would be/foo/bar/baz.py. path need not exist, however if it does not conformto PEP 3147 format, a ValueError is raised. Ifsys.implementation.cache_tag is not defined,NotImplementedError is raised.

在 3.3 版更改: Raise NotImplementedError whensys.implementation.cache_tag is not defined.

3.4 版后已移除: Use importlib.util.source_from_cache() instead.

  • imp.get_tag()
  • Return the PEP 3147 magic tag string matching this version of Python'smagic number, as returned by get_magic().

3.4 版后已移除: Use sys.implementation.cache_tag directly startingin Python 3.3.

The following functions help interact with the import system's internallocking mechanism. Locking semantics of imports are an implementationdetail which may vary from release to release. However, Python ensuresthat circular imports work without any deadlocks.

  • imp.lock_held()
  • Return True if the global import lock is currently held, elseFalse. On platforms without threads, always return False.

On platforms with threads, a thread executing an import first holds aglobal import lock, then sets up a per-module lock for the rest of theimport. This blocks other threads from importing the same module untilthe original import completes, preventing other threads from seeingincomplete module objects constructed by the original thread. Anexception is made for circular imports, which by construction have toexpose an incomplete module object at some point.

在 3.3 版更改: The locking scheme has changed to per-module locks forthe most part. A global import lock is kept for some critical tasks,such as initializing the per-module locks.

3.4 版后已移除.

  • imp.acquire_lock()
  • Acquire the interpreter's global import lock for the current thread.This lock should be used by import hooks to ensure thread-safety whenimporting modules.

Once a thread has acquired the import lock, the same thread may acquire itagain without blocking; the thread must release it once for each time it hasacquired it.

On platforms without threads, this function does nothing.

在 3.3 版更改: The locking scheme has changed to per-module locks forthe most part. A global import lock is kept for some critical tasks,such as initializing the per-module locks.

3.4 版后已移除.

  • imp.release_lock()
  • Release the interpreter's global import lock. On platforms withoutthreads, this function does nothing.

在 3.3 版更改: The locking scheme has changed to per-module locks forthe most part. A global import lock is kept for some critical tasks,such as initializing the per-module locks.

3.4 版后已移除.

The following constants with integer values, defined in this module, are usedto indicate the search result of find_module().

  • imp.PY_SOURCE
  • The module was found as a source file.

3.3 版后已移除.

  • imp.PY_COMPILED
  • The module was found as a compiled code object file.

3.3 版后已移除.

  • imp.C_EXTENSION
  • The module was found as dynamically loadable shared library.

3.3 版后已移除.

  • imp.PKG_DIRECTORY
  • The module was found as a package directory.

3.3 版后已移除.

  • imp.C_BUILTIN
  • The module was found as a built-in module.

3.3 版后已移除.

  • imp.PY_FROZEN
  • The module was found as a frozen module.

3.3 版后已移除.

  • class imp.NullImporter(path_string)
  • The NullImporter type is a PEP 302 import hook that handlesnon-directory path strings by failing to find any modules. Calling this typewith an existing directory or empty string raises ImportError.Otherwise, a NullImporter instance is returned.

Instances have only one method:

  • findmodule(_fullname[, path])
  • This method always returns None, indicating that the requested module couldnot be found.

在 3.3 版更改: None is inserted into sys.path_importer_cache instead of aninstance of NullImporter.

3.4 版后已移除: Insert None into sys.path_importer_cache instead.

示例

The following function emulates what was the standard import statement up toPython 1.4 (no hierarchical module names). (This implementation wouldn't workin that version, since find_module() has been extended andload_module() has been added in 1.4.)

  1. import imp
  2. import sys
  3.  
  4. def __import__(name, globals=None, locals=None, fromlist=None):
  5. # Fast path: see if the module has already been imported.
  6. try:
  7. return sys.modules[name]
  8. except KeyError:
  9. pass
  10.  
  11. # If any of the following calls raises an exception,
  12. # there's a problem we can't handle -- let the caller handle it.
  13.  
  14. fp, pathname, description = imp.find_module(name)
  15.  
  16. try:
  17. return imp.load_module(name, fp, pathname, description)
  18. finally:
  19. # Since we may exit via an exception, close fp explicitly.
  20. if fp:
  21. fp.close()