Issue #13959: Introduce importlib.find_loader().

The long-term goal is to deprecate imp.find_module() in favour of this
API, but it will take some time as some APIs explicitly return/use what
imp.find_module() returns.

diff --git a/Doc/library/importlib.rst b/Doc/library/importlib.rst
index cac5251e0816050822fa2c4b20f72d41a8f342a3..0bc1b65a4b6a1e69fb4416fc30b88bb7494af2be 100644 (file)
--- a/Doc/library/importlib.rst
+++ b/Doc/library/importlib.rst
@@ -86,6 +86,20 @@ Functions
     that was imported (e.g. ``pkg.mod``), while :func:`__import__` returns the
     top-level package or module (e.g. ``pkg``).
 
+.. function:: find_loader(name, path=None)
+
+   Find the loader for a module, optionally within the specified *path*. If the
+   module is in :attr:`sys.modules`, then ``sys.modules[name].__loader__`` is
+   returned (unless the loader would be ``None``, in which case
+   :exc:`ValueError` is raised). Otherwise a search using :attr:`sys.meta_path`
+   is done. ``None`` is returned if no loader is found.
+
+   A dotted name does not have its parent's implicitly imported. If that is
+   desired (although not nessarily required to find the loader, it will most
+   likely be needed if the loader actually is used to load the module), then
+   you will have to import the packages containing the module prior to calling
+   this function.
+
 .. function:: invalidate_caches()
 
    Invalidate the internal caches of the finders stored at

diff --git a/Lib/importlib/__init__.py b/Lib/importlib/__init__.py
index 57fb284ab12a85d6ba67eeafc36d64f57cd1b0a6..90e163c433b0889812993279dbaded27eab1077c 100644 (file)
--- a/Lib/importlib/__init__.py
+++ b/Lib/importlib/__init__.py
@@ -29,6 +29,30 @@ def invalidate_caches():
             finder.invalidate_caches()
 
 
+def find_loader(name, path=None):
+    """Find the loader for the specified module.
+
+    First, sys.modules is checked to see if the module was already imported. If
+    so, then sys.modules[name].__loader__ is returned. If that happens to be
+    set to None, then ValueError is raised. If the module is not in
+    sys.modules, then sys.meta_path is searched for a suitable loader with the
+    value of 'path' given to the finders. None is returned if no loader could
+    be found.
+
+    Dotted names do not have their parent packages implicitly imported.
+
+    """
+    try:
+        loader = sys.modules[name].__loader__
+        if loader is None:
+            raise ValueError('{}.__loader__ is None'.format(name))
+        else:
+            return loader
+    except KeyError:
+        pass
+    return _bootstrap._find_module(name, path)
+
+
 def import_module(name, package=None):
     """Import a module.
 

diff --git a/Lib/importlib/test/test_api.py b/Lib/importlib/test/test_api.py
index cc147c200bd6110739919eb1661ce8c150a0d4c1..b7d6cb4effe186c3eda064b37b08acdb02f4da1f 100644 (file)
--- a/Lib/importlib/test/test_api.py
+++ b/Lib/importlib/test/test_api.py
@@ -85,6 +85,54 @@ class ImportModuleTests(unittest.TestCase):
         self.assertEqual(b_load_count, 1)
 
 
+class FindLoaderTests(unittest.TestCase):
+
+    class FakeMetaFinder:
+        @staticmethod
+        def find_module(name, path=None): return name, path
+
+    def test_sys_modules(self):
+        # If a module with __loader__ is in sys.modules, then return it.
+        name = 'some_mod'
+        with util.uncache(name):
+            module = imp.new_module(name)
+            loader = 'a loader!'
+            module.__loader__ = loader
+            sys.modules[name] = module
+            found = importlib.find_loader(name)
+            self.assertEqual(loader, found)
+
+    def test_sys_modules_loader_is_None(self):
+        # If sys.modules[name].__loader__ is None, raise ValueError.
+        name = 'some_mod'
+        with util.uncache(name):
+            module = imp.new_module(name)
+            module.__loader__ = None
+            sys.modules[name] = module
+            with self.assertRaises(ValueError):
+                importlib.find_loader(name)
+
+    def test_success(self):
+        # Return the loader found on sys.meta_path.
+        name = 'some_mod'
+        with util.uncache(name):
+            with util.import_state(meta_path=[self.FakeMetaFinder]):
+                self.assertEqual((name, None), importlib.find_loader(name))
+
+    def test_success_path(self):
+        # Searching on a path should work.
+        name = 'some_mod'
+        path = 'path to some place'
+        with util.uncache(name):
+            with util.import_state(meta_path=[self.FakeMetaFinder]):
+                self.assertEqual((name, path),
+                                 importlib.find_loader(name, path))
+
+    def test_nothing(self):
+        # None is returned upon failure to find a loader.
+        self.assertIsNone(importlib.find_loader('nevergoingtofindthismodule'))
+
+
 class InvalidateCacheTests(unittest.TestCase):
 
     def test_method_called(self):
@@ -114,7 +162,7 @@ class InvalidateCacheTests(unittest.TestCase):
 
 def test_main():
     from test.support import run_unittest
-    run_unittest(ImportModuleTests)
+    run_unittest(ImportModuleTests, FindLoaderTests, InvalidateCacheTests)
 
 
 if __name__ == '__main__':

diff --git a/Lib/pyclbr.py b/Lib/pyclbr.py
index 52cbdd53ab3d106ec53496079c8fd924526f4bc6..4cd85b9a80953632347843190ed5b0c7f6cc82ca 100644 (file)
--- a/Lib/pyclbr.py
+++ b/Lib/pyclbr.py
@@ -39,8 +39,10 @@ Instances of this class have the following instance variables:
         lineno -- the line in the file on which the class statement occurred
 """
 
+import io
+import os
 import sys
-import imp
+import importlib
 import tokenize
 from token import NAME, DEDENT, OP
 from operator import itemgetter
@@ -133,19 +135,24 @@ def _readmodule(module, path, inpackage=None):
     # Search the path for the module
     f = None
     if inpackage is not None:
-        f, fname, (_s, _m, ty) = imp.find_module(module, path)
+        search_path = path
     else:
-        f, fname, (_s, _m, ty) = imp.find_module(module, path + sys.path)
-    if ty == imp.PKG_DIRECTORY:
-        dict['__path__'] = [fname]
-        path = [fname] + path
-        f, fname, (_s, _m, ty) = imp.find_module('__init__', [fname])
+        search_path = path + sys.path
+    loader = importlib.find_loader(fullmodule, search_path)
+    fname = loader.get_filename(fullmodule)
     _modules[fullmodule] = dict
-    if ty != imp.PY_SOURCE:
+    if loader.is_package(fullmodule):
+        dict['__path__'] = [os.path.dirname(fname)]
+    try:
+        source = loader.get_source(fullmodule)
+        if source is None:
+            return dict
+    except (AttributeError, ImportError):
         # not Python source, can't do anything with this module
-        f.close()
         return dict
 
+    f = io.StringIO(source)
+
     stack = [] # stack of (class, indent) pairs
 
     g = tokenize.generate_tokens(f.readline)

diff --git a/Misc/NEWS b/Misc/NEWS
index 78a5ef9e5773dd1205739f3246e6d571a7f18021..a7d1b4492223a04788a4b08f3b471511662e8d61 100644 (file)
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -23,6 +23,8 @@ Core and Builtins
 Library
 -------
 
+- Issue #13959: Introduce importlib.find_loader().
+
 - Issue #14082: shutil.copy2() now copies extended attributes, if possible.
   Patch by Hynek Schlawack.