Commit Graph

49 Commits

Author SHA1 Message Date
Brett Cannon 461c813164 Issue #15111: When a module was imported using a 'from import'
statement (e.g. ``from distutils import msvc9compiler``) that triggers
an ImportError of its own (e.g. the non-existence of winreg), let that
exception propagate instead of raising a generic ImportError for the
module being requested (e.g. msvc9compiler).
2012-07-10 10:05:00 -04:00
Eric V. Smith b10951549b Use assertIsNone. Thanks Terry Reedy. 2012-06-28 06:15:01 -04:00
Eric V. Smith faae3adbb9 Changed importlib tests to use assertIs, assertIsInstance, etc., instead of just assertTrue. 2012-06-27 15:26:26 -04:00
Antoine Pitrou 6efa50a384 Issue #14583: Fix importlib bug when a package's __init__.py would first import one of its modules then raise an error. 2012-05-07 21:41:59 +02:00
Brett Cannon aa93642a35 Issue #14605: Use None in sys.path_importer_cache to represent no
finder instead of using some (now non-existent) implicit finder.
2012-04-27 15:30:58 -04:00
Brett Cannon ce418b448f Issue #14605: Stop having implicit entries for sys.meta_path.
ImportWarning is raised if sys.meta_path is found to be empty.
2012-04-27 14:01:58 -04:00
Brett Cannon e0d88a173c Issue #14605: Make explicit the entries on sys.path_hooks that used to
be implicit.

Added a warning for when sys.path_hooks is found to be empty. Also
changed the meaning of None in sys.path_importer_cache to represent
trying sys.path_hooks again (an interpretation of previous semantics).
Also added a warning for when None was found.

The long-term goal is for None in sys.path_importer_cache to represent
the same as imp.NullImporter: no finder found for that sys.path entry.
2012-04-25 20:54:04 -04:00
Benjamin Peterson d76bc7abac rollback 005fd1fe31ab (see #14609 and #14582)
Being able to overload a sys.module entry during import of a module was broken
by this changeset.
2012-04-18 10:55:43 -04:00
Brett Cannon 7bd329d800 Issue #12599: Be more strict in accepting None vs. a false-like object
in importlib.

Thanks to PJE for pointing out the issue and Nick Coghlan for filing
the bug.
2012-04-17 21:41:35 -04:00
Brett Cannon 273323cf68 Issue #14592: A relative import will raise a KeyError if __package__
or __name__ are not set in globals.

Thanks to Stefan Behnel for the bug report.
2012-04-17 19:05:11 -04:00
Brett Cannon 881535b726 Issue #14582: Import returns the module returned by a loader instead
of sys.modules when possible.

This is being done for two reasons. One is to gain a little bit of
performance by skipping an unnecessary dict lookup in sys.modules. But
the other (and main) reason is to be a little bit more clear in how
things should work from the perspective of import's interactions with
loaders. Otherwise loaders can easily forget to return the module even
though PEP 302 explicitly states they are expected to return the module
they loaded.
2012-04-15 15:24:04 -04:00
Brett Cannon 49f8d8b016 Handle importing pkg.mod by executing
__import__('mod', {'__packaging__': 'pkg', level=1) w/o properly (and
thus not segfaulting).
2012-04-14 21:50:00 -04:00
Brett Cannon fd0741555b Issue #2377: Make importlib the implementation of __import__().
importlib._bootstrap is now frozen into Python/importlib.h and stored
as _frozen_importlib in sys.modules. Py_Initialize() loads the frozen
code along with sys and imp and then uses _frozen_importlib._install()
to set builtins.__import__() w/ _frozen_importlib.__import__().
2012-04-14 14:10:13 -04:00
Brett Cannon bbb6680ee5 Have importlib take advantage of ImportError's new 'name' and 'path'
attributes.
2012-04-12 21:09:01 -04:00
Brett Cannon d62cd5627f Issue #14500: Fix importlib.test.import_.test_packages to clean up
after itself properly.
2012-04-06 13:13:08 -04:00
Brett Cannon 927d87470a If a module injects something into sys.modules as a side-effect of
importation, then respect that injection.

Discovered thanks to Lib/xml/parsers/expat.py injecting
xml.parsers.expat.errors and etree now importing that directly as a
module.
2012-04-02 20:33:56 -04:00
Brett Cannon f500778f65 Improper type for __package__ should raise TypeError, not ValueError. 2012-02-23 18:29:12 -05:00
Antoine Pitrou c541f8ef40 Issue #14043: Speed up importlib's _FileFinder by at least 8x, and add a new importlib.invalidate_caches() function.
importlib is now often faster than imp.find_module() at finding modules.
2012-02-20 01:48:16 +01:00
Brett Cannon f58d45c649 Tweak the handling of the empty string in sys.path for importlib.
It seems better to cache the finder for the cwd under its full path
insetad of '' in case the cwd changes. Otherwise FileFinder needs to
dynamically change itself based on whether it is given '' instead of
caching a finder for every change to the cwd.
2012-02-16 18:12:00 -05:00
Brett Cannon 3b1a06c1ea importlib.__import__() now raises ValueError when level < 0.
This is to bring it more in line with what PEP 328 set out to do with
removing ambiguous absolute/relative import semantics.
2012-02-16 17:47:48 -05:00
Brett Cannon b4e63b3177 Use the cwd when the empty string is found in sys.path. This leads to
__file__ being an absolute path when the module is found in the
current directory.
2012-02-08 18:52:56 -05:00
Ezio Melotti b3aedd4862 #9424: Replace deprecated assert* methods in the Python test suite. 2010-11-20 19:04:17 +00:00
Brett Cannon 61b14251d3 Make importlib.abc.SourceLoader the primary mechanism for importlib.
This required moving the class from importlib/abc.py into
importlib/_bootstrap.py and jiggering some code to work better with the class.
This included changing how the file finder worked to better meet import
semantics. This also led to fixing importlib to handle the empty string from
sys.path as import currently does (and making me wish we didn't support that
instead just required people to insert '.' instead to represent cwd).

It also required making the new set_data abstractmethod create
any needed subdirectories implicitly thanks to __pycache__ (it was either this
or grow the SourceLoader ABC to gain an 'exists' method and either a mkdir
method or have set_data with no data arg mean to create a directory).

Lastly, as an optimization the file loaders cache the file path where the
finder found something to use for loading (this is thanks to having a
sourceless loader separate from the source loader to simplify the code and
cut out stat calls).
Unfortunately test_runpy assumed a loader would always work for a module, even
if you changed from underneath it what it was expected to work with. By simply
dropping the previous loader in test_runpy so the proper loader can be returned
by the finder fixed the failure.

At this point importlib deviates from import on two points:

1. The exception raised when trying to import a file is different (import does
an explicit file check to print a special message, importlib just says the path
cannot be imported as if it was just some module name).

2. the co_filename on a code object is not being set to where bytecode was
actually loaded from instead of where the marshalled code object originally
came from (a solution for this has already been agreed upon on python-dev but has
not been implemented yet; issue8611).
2010-07-03 21:48:25 +00:00
Brett Cannon 1c1dcbfd5d Trying to import a submodule from another module and not a package was raising
AttributeError in importlib when it should be an ImportError.

Found when running importlib against test_runpy.
2009-08-30 20:22:21 +00:00
Brett Cannon de4ebfe559 When the globals argument to importlib.__import__() contained any value for
__package__, it was used. This was incorrect since it could be set to None to
represent the fact that a proper value was unknown. Now None will trigger the
calculation for __package__.

Discovered when running importlib against test_importhooks.
2009-08-30 19:53:48 +00:00
Brett Cannon 6afbaef2fd Raise TypeError if the name given to importlib.__import__() lacks an rpartition
attribute. Was throwing AttributeError before. Discovered when running
test_builtin against importlib.

This exception change is specific to importlib.__import__() and does not apply to
import_module() as it is being done for compatibility reasons only.
2009-08-30 19:08:58 +00:00
Brett Cannon 44b28a9f32 Fix the importlib_only test decorator to work again; don't capture the flag variable as it might change later. 2009-08-30 18:59:21 +00:00
Brett Cannon a6503606f9 Use the public API, not a private one. 2009-08-30 18:40:23 +00:00
Brett Cannon 9e0e1a63c8 Allow importlib.__import__ to accept any iterable for fromlist. Discovered when
running importlib against test___all__.
2009-08-30 18:28:46 +00:00
Brett Cannon 4d75fc1ce9 Have importlib raise ImportError if None is found in sys.modules. This matches
current import semantics.
2009-08-30 03:47:36 +00:00
Brett Cannon 2153dc001f Move over to using assertRaises as a context manager for importlib tests.
Obviously one shouldn't do whole sale conversions like this, but I was already
going through the test code and I was bored at the airport.
2009-08-27 23:49:21 +00:00
Brett Cannon c5951fc996 Make __package__ setting tests specific to importlib. Also move to assertRaises context manager. 2009-08-27 23:46:38 +00:00
Brett Cannon 6d2bb7f5aa Move a test-skipping decorator over to unittest.skipIf. 2009-08-27 23:44:18 +00:00
R. David Murray 9b2611ee40 Remove unused import for function that no longer exists. 2009-07-01 02:51:00 +00:00
Benjamin Peterson c9c0f201fe convert old fail* assertions to assert* 2009-06-30 23:06:06 +00:00
Raymond Hettinger 554290d920 Fixup/simplify another nested context manager. 2009-06-12 11:25:59 +00:00
Brett Cannon 8593a75688 Fix importlib.machinery.PathFinder.find_module() to essentially skip over None
entries in sys.path_importer_cache. While this differs from semantics in how
__import__ works, it prevents any implicit semantics from taking hold with
users.
2009-03-30 19:57:15 +00:00
Brett Cannon 7aa21f75c1 A few more docstring/API cleanups for importlib. 2009-03-15 00:53:05 +00:00
Brett Cannon ba3fcf17c8 Remove a dead XXX comment. 2009-03-09 01:58:13 +00:00
Brett Cannon 2dee597e05 Do some cleanup in importlib:
+ Ditch using arguments to super().
+ Ditch subclassing from object directly.
+ Move directory check out of chaining path hook to file path hook/finder.
+ Rename some classes to better reflect they are finders, not importers.
2009-02-21 03:15:37 +00:00
Brett Cannon 32732e3fbd Change importlib.machinery.PathFinder to not have implicit semantics (that's
not handled by importlib._bootstrap._DefaultPathFinder).
2009-02-15 05:48:13 +00:00
Brett Cannon 2c318a1390 Rewrite the code implementing __import__ for importlib. Now it is much simpler
and relies much more on meta path finders to abstract out various parts of
import.

As part of this the semantics for import_module tightened up and now follow
__import__ much more closely (biggest thing is that the 'package' argument must
now already be imported, else a SystemError is raised).
2009-02-07 01:15:27 +00:00
Brett Cannon 6411aa5dd4 Finish implementing tests for importlib.machinery.PathFinder by testing that
implicit hooks are handled properly.
2009-02-06 00:07:49 +00:00
Brett Cannon 1f9bcd38a9 Add tests for using sys.path_hooks by importlib.machinery.PathFinder. 2009-02-05 23:36:02 +00:00
Brett Cannon 48f5c4eec6 Implement test for sys.path_importer_cache having None for
importlib.machinery.PathFinder.
2009-02-05 22:02:03 +00:00
Brett Cannon 939ea76b3c Begin writing tests for importlib.machinery.PathFinder. 2009-02-05 02:53:23 +00:00
Brett Cannon d720b36248 Move import semantic util code to importlib.test.import_.util. 2009-02-01 04:28:04 +00:00
Brett Cannon bcb26c53c0 Rename importlib.test.support to importlib.test.util. 2009-02-01 04:00:05 +00:00
Brett Cannon 23cbd8a656 Add initial implementation of importlib. See the NOTES files for what is
planned for the package.

There are no docs yet, but they are coming once the API for the first new
function, importlib.import_module() is finalized.
2009-01-18 00:24:28 +00:00