PEP 302 + zipimport:

- new import hooks in import.c, exposed in the sys module
- new module called 'zipimport'
- various changes to allow bootstrapping from zip files

I hope I didn't break the Windows build (or anything else for that
matter), but then again, it's been sitting on sf long enough...

Regarding the latest discussions on python-dev: zipimport sets
pkg.__path__ as specified in PEP 273, and likewise, sys.path item such as
/path/to/Archive.zip/subdir/ are supported again.

3 files changed, 388 insertions, 9 deletions

diff --git a/Lib/site.py b/Lib/site.py
index a672765..0ab7a0f 100644
--- a/Lib/site.py
+++ b/Lib/site.py
@@ -73,16 +73,11 @@ del m
 # only absolute pathnames, even if we're running from the build directory.
 L = []
 _dirs_in_sys_path = {}
+dir = dircase = None  # sys.path may be empty at this point
 for dir in sys.path:
-    # Filter out paths that don't exist, but leave in the empty string
-    # since it's a special case. We also need to special-case the Mac,
-    # as file names are allowed on sys.path there.
-    if sys.platform != 'mac':
-        if dir and not os.path.isdir(dir):
-            continue
-    else:
-        if dir and not os.path.exists(dir):
-            continue
+    # Filter out duplicate paths (on case-insensitive file systems also
+    # if they only differ in case); turn relative paths into absolute
+    # paths.
     dir, dircase = makepath(dir)
     if not dircase in _dirs_in_sys_path:
         L.append(dir)
diff --git a/Lib/test/test_importhooks.py b/Lib/test/test_importhooks.py
new file mode 100644
index 0000000..070b578
--- /dev/null
+++ b/Lib/test/test_importhooks.py
@@ -0,0 +1,204 @@
+import sys
+import imp
+import os
+import unittest
+from test import test_support
+
+
+test_src = """\
+def get_name():
+    return __name__
+def get_file():
+    return __file__
+"""
+
+test_co = compile(test_src, "<???>", "exec")
+test_path = "!!!_test_!!!"
+
+
+class ImportTracker:
+    """Importer that only tracks attempted imports."""
+    def __init__(self):
+        self.imports = []
+    def find_module(self, fullname, path=None):
+        self.imports.append(fullname)
+        return None
+
+
+class TestImporter:
+
+    modules = {
+        "hooktestmodule": (False, test_co),
+        "hooktestpackage": (True, test_co),
+        "hooktestpackage.sub": (True, test_co),
+        "hooktestpackage.sub.subber": (False, test_co),
+    }
+
+    def __init__(self, path=test_path):
+        if path != test_path:
+            # if out class is on sys.path_hooks, we must raise
+            # ImportError for any path item that we can't handle.
+            raise ImportError
+        self.path = path
+
+    def _get__path__(self):
+        raise NotImplementedError
+
+    def find_module(self, fullname, path=None):
+        if fullname in self.modules:
+            return self
+        else:
+            return None
+
+    def load_module(self, fullname):
+        ispkg, code = self.modules[fullname]
+        mod = imp.new_module(fullname)
+        sys.modules[fullname] = mod
+        mod.__file__ = "<%s>" % self.__class__.__name__
+        mod.__loader__ = self
+        if ispkg:
+            mod.__path__ = self._get__path__()
+        exec code in mod.__dict__
+        return mod
+
+
+class MetaImporter(TestImporter):
+    def _get__path__(self):
+        return []
+
+class PathImporter(TestImporter):
+    def _get__path__(self):
+        return [self.path]
+
+
+class ImportBlocker:
+    """Place an ImportBlocker instance on sys.meta_path and you
+    can be sure the modules you specified can't be imported, even
+    if it's a builtin."""
+    def __init__(self, *namestoblock):
+        self.namestoblock = dict.fromkeys(namestoblock)
+    def find_module(self, fullname, path=None):
+        if fullname in self.namestoblock:
+            return self
+        return None
+    def load_module(self, fullname):
+        raise ImportError, "I dare you"
+
+
+class ImpWrapper:
+
+    def __init__(self, path=None):
+        if path is not None and not os.path.isdir(path):
+            raise ImportError
+        self.path = path
+
+    def find_module(self, fullname, path=None):
+        subname = fullname.split(".")[-1]
+        if subname != fullname and self.path is None:
+            return None
+        if self.path is None:
+            path = None
+        else:
+            path = [self.path]
+        try:
+            file, filename, stuff = imp.find_module(subname, path)
+        except ImportError:
+            return None
+        return ImpLoader(file, filename, stuff)
+
+
+class ImpLoader:
+
+    def __init__(self, file, filename, stuff):
+        self.file = file
+        self.filename = filename
+        self.stuff = stuff
+
+    def load_module(self, fullname):
+        mod = imp.load_module(fullname, self.file, self.filename, self.stuff)
+        if self.file:
+            self.file.close()
+        mod.__loader__ = self  # for introspection
+        return mod
+
+
+class ImportHooksBaseTestCase(unittest.TestCase):
+
+    def setUp(self):
+        self.path = sys.path[:]
+        self.meta_path = sys.meta_path[:]
+        self.path_hooks = sys.path_hooks[:]
+        sys.path_importer_cache.clear()
+        self.tracker = ImportTracker()
+        sys.meta_path.insert(0, self.tracker)
+
+    def tearDown(self):
+        sys.path[:] = self.path
+        sys.meta_path[:] = self.meta_path
+        sys.path_hooks[:] = self.path_hooks
+        sys.path_importer_cache.clear()
+        for fullname in self.tracker.imports:
+            if fullname in sys.modules:
+                del sys.modules[fullname]
+
+
+class ImportHooksTestCase(ImportHooksBaseTestCase):
+
+    def doTestImports(self, importer=None):
+        import hooktestmodule
+        import hooktestpackage
+        import hooktestpackage.sub
+        import hooktestpackage.sub.subber
+        self.assertEqual(hooktestmodule.get_name(),
+                         "hooktestmodule")
+        self.assertEqual(hooktestpackage.get_name(),
+                         "hooktestpackage")
+        self.assertEqual(hooktestpackage.sub.get_name(),
+                         "hooktestpackage.sub")
+        self.assertEqual(hooktestpackage.sub.subber.get_name(),
+                         "hooktestpackage.sub.subber")
+        if importer:
+            self.assertEqual(hooktestmodule.__loader__, importer)
+            self.assertEqual(hooktestpackage.__loader__, importer)
+            self.assertEqual(hooktestpackage.sub.__loader__, importer)
+            self.assertEqual(hooktestpackage.sub.subber.__loader__, importer)
+
+    def testMetaPath(self):
+        i = MetaImporter()
+        sys.meta_path.append(i)
+        self.doTestImports(i)
+
+    def testPathHook(self):
+        sys.path_hooks.append(PathImporter)
+        sys.path.append(test_path)
+        self.doTestImports()
+
+    def testBlocker(self):
+        mname = "exceptions"  # an arbitrary harmless builtin module
+        if mname in sys.modules:
+            del sys.modules[mname]
+        sys.meta_path.append(ImportBlocker(mname))
+        try:
+            __import__(mname)
+        except ImportError:
+            pass
+        else:
+            self.fail("'%s' was not supposed to be importable" % mname)
+
+    def testImpWrapper(self):
+        i = ImpWrapper()
+        sys.meta_path.append(i)
+        sys.path_hooks.append(ImpWrapper)
+        mnames = ("colorsys", "urlparse", "distutils.core", "compiler.misc")
+        for mname in mnames:
+            parent = mname.split(".")[0]
+            for n in sys.modules.keys():
+                if n.startswith(parent):
+                    del sys.modules[n]
+        for mname in mnames:
+            m = __import__(mname, globals(), locals(), ["__dummy__"])
+            m.__loader__  # to make sure we actually handled the import
+
+
+if __name__ == "__main__":
+    test_support.run_unittest(ImportHooksTestCase)
diff --git a/Lib/test/test_zipimport.py b/Lib/test/test_zipimport.py
new file mode 100644
index 0000000..07520a7
--- /dev/null
+++ b/Lib/test/test_zipimport.py
@@ -0,0 +1,180 @@
+import sys
+import os
+import marshal
+import imp
+import struct
+import time
+
+import zlib # implied prerequisite
+from zipfile import ZipFile, ZipInfo, ZIP_STORED, ZIP_DEFLATED
+from test import test_support
+from test.test_importhooks import ImportHooksBaseTestCase, test_src, test_co
+
+import zipimport
+
+
+def make_pyc(co, mtime):
+    data = marshal.dumps(co)
+    pyc = imp.get_magic() + struct.pack("<i", mtime) + data
+    return pyc
+
+NOW = time.time()
+test_pyc = make_pyc(test_co, NOW)
+
+
+if __debug__:
+    pyc_ext = ".pyc"
+else:
+    pyc_ext = ".pyo"
+
+
+TESTMOD = "ziptestmodule"
+TESTPACK = "ziptestpackage"
+TEMP_ZIP = "junk95142.zip"
+
+
+class UncompressedZipImportTestCase(ImportHooksBaseTestCase):
+
+    compression = ZIP_STORED
+
+    def setUp(self):
+        # We're reusing the zip archive path, so we must clear the
+        # cached directory info.
+        zipimport._zip_directory_cache.clear()
+        ImportHooksBaseTestCase.setUp(self)
+
+    def doTest(self, expected_ext, files, *modules):
+        z = ZipFile(TEMP_ZIP, "w")
+        try:
+            for name, (mtime, data) in files.items():
+                zinfo = ZipInfo(name, time.localtime(mtime))
+                zinfo.compress_type = self.compression
+                z.writestr(zinfo, data)
+            z.close()
+            sys.path.insert(0, TEMP_ZIP)
+
+            mod = __import__(".".join(modules), globals(), locals(),
+                             ["__dummy__"])
+            file = mod.get_file()
+            self.assertEquals(file, os.path.join(TEMP_ZIP,
+                              os.sep.join(modules) + expected_ext))
+        finally:
+            z.close()
+            os.remove(TEMP_ZIP)
+
+    def testAFakeZlib(self):
+        #
+        # This could cause a stack overflow before: importing zlib.py
+        # from a compressed archive would cause zlib to be imported
+        # which would find zlib.py in the archive, which would... etc.
+        #
+        # This test *must* be executed first: it must be the first one
+        # to trigger zipimport to import zlib (zipimport caches the
+        # zlib.decompress function object, after which the problem being
+        # tested here wouldn't be a problem anymore...
+        # (Hence the 'A' in the test method name: to make it the first
+        # item in a list sorted by name, like unittest.makeSuite() does.)
+        #
+        if "zlib" in sys.modules:
+            del sys.modules["zlib"]
+        files = {"zlib.py": (NOW, test_src)}
+        try:
+            self.doTest(".py", files, "zlib")
+        except ImportError:
+            if self.compression != ZIP_DEFLATED:
+                self.fail("expected test to not raise ImportError")
+        else:
+            if self.compression != ZIP_STORED:
+                self.fail("expected test to raise ImportError")
+
+    def testPy(self):
+        files = {TESTMOD + ".py": (NOW, test_src)}
+        self.doTest(".py", files, TESTMOD)
+
+    def testPyc(self):
+        files = {TESTMOD + pyc_ext: (NOW, test_pyc)}
+        self.doTest(pyc_ext, files, TESTMOD)
+
+    def testBoth(self):
+        files = {TESTMOD + ".py": (NOW, test_src),
+                 TESTMOD + pyc_ext: (NOW, test_pyc)}
+        self.doTest(pyc_ext, files, TESTMOD)
+
+    def testBadMagic(self):
+        # make pyc magic word invalid, forcing loading from .py
+        m0 = ord(test_pyc[0])
+        m0 ^= 0x04  # flip an arbitrary bit
+        badmagic_pyc = chr(m0) + test_pyc[1:]
+        files = {TESTMOD + ".py": (NOW, test_src),
+                 TESTMOD + pyc_ext: (NOW, badmagic_pyc)}
+        self.doTest(".py", files, TESTMOD)
+
+    def testBadMagic2(self):
+        # make pyc magic word invalid, causing an ImportError
+        m0 = ord(test_pyc[0])
+        m0 ^= 0x04  # flip an arbitrary bit
+        badmagic_pyc = chr(m0) + test_pyc[1:]
+        files = {TESTMOD + pyc_ext: (NOW, badmagic_pyc)}
+        try:
+            self.doTest(".py", files, TESTMOD)
+        except ImportError:
+            pass
+        else:
+            self.fail("expected ImportError; import from bad pyc")
+
+    def testBadMTime(self):
+        t3 = ord(test_pyc[7])
+        t3 ^= 0x02  # flip the second bit -- not the first as that one
+                    # isn't stored in the .py's mtime in the zip archive.
+        badtime_pyc = test_pyc[:7] + chr(t3) + test_pyc[8:]
+        files = {TESTMOD + ".py": (NOW, test_src),
+                 TESTMOD + pyc_ext: (NOW, badtime_pyc)}
+        self.doTest(".py", files, TESTMOD)
+
+    def testPackage(self):
+        packdir = TESTPACK + os.sep
+        files = {packdir + "__init__" + pyc_ext: (NOW, test_pyc),
+                 packdir + TESTMOD + pyc_ext: (NOW, test_pyc)}
+        self.doTest(pyc_ext, files, TESTPACK, TESTMOD)
+
+    def testDeepPackage(self):
+        packdir = TESTPACK + os.sep
+        packdir2 = packdir + packdir
+        files = {packdir + "__init__" + pyc_ext: (NOW, test_pyc),
+                 packdir2 + "__init__" + pyc_ext: (NOW, test_pyc),
+                 packdir2 + TESTMOD + pyc_ext: (NOW, test_pyc)}
+        self.doTest(pyc_ext, files, TESTPACK, TESTPACK, TESTMOD)
+
+    def testGetData(self):
+        z = ZipFile(TEMP_ZIP, "w")
+        z.compression = self.compression
+        try:
+            name = "testdata.dat"
+            data = "".join([chr(x) for x in range(256)]) * 500
+            z.writestr(name, data)
+            z.close()
+            zi = zipimport.zipimporter(TEMP_ZIP)
+            self.assertEquals(data, zi.get_data(name))
+        finally:
+            z.close()
+            os.remove(TEMP_ZIP)
+
+    def testImporterAttr(self):
+        src = """if 1:  # indent hack
+        def get_file():
+            return __file__
+        if __importer__.get_data("some.data") != "some data":
+            raise AssertionError, "bad data"\n"""
+        pyc = make_pyc(compile(src, "<???>", "exec"), NOW)
+        files = {TESTMOD + pyc_ext: (NOW, pyc),
+                 "some.data": (NOW, "some data")}
+        self.doTest(pyc_ext, files, TESTMOD)
+
+
+class CompressedZipImportTestCase(UncompressedZipImportTestCase):
+    compression = ZIP_DEFLATED
+
+
+if __name__ == "__main__":
+    test_support.run_unittest(UncompressedZipImportTestCase)
+    test_support.run_unittest(CompressedZipImportTestCase)