From 265edf6e1fbe92ae4f1d7d753add389ae76f813f Mon Sep 17 00:00:00 2001 From: Harjoth Khara Date: Sat, 10 Oct 2026 12:56:51 -0700 Subject: [PATCH] [3.14] gh-141276: Avoid compiling zipimport source in get_filename (GH-153799) `zipimporter.get_filename()` compiled source modules just to work out the module's file name. `find_spec()` calls it through `spec_from_loader()`, so importing a source module from a ZIP archive compiled it twice and emitted every `SyntaxWarning` twice. `get_filename()` now returns before reading or compiling the source. As a result, a `SyntaxError` in a zipped source module is now raised by `exec_module()` rather than by `find_spec()` or `get_filename()`, and the source isn't compiled twice during import, which matches the filesystem-based loaders. (cherry picked from commit 230917f5072c5c6b9373374a84d52935d7afa3b5) --- Doc/library/zipimport.rst | 2 +- Lib/test/test_zipimport.py | 75 +++++++++++++++++++ Lib/zipimport.py | 14 ++-- ...-07-15-21-45-33.gh-issue-141276.xH7qP2.rst | 6 ++ 4 files changed, 90 insertions(+), 7 deletions(-) create mode 100644 Misc/NEWS.d/next/Library/2026-07-15-21-45-33.gh-issue-141276.xH7qP2.rst diff --git a/Doc/library/zipimport.rst b/Doc/library/zipimport.rst index 7c33db6848aec8f..179e4c119ab0856 100644 --- a/Doc/library/zipimport.rst +++ b/Doc/library/zipimport.rst @@ -123,7 +123,7 @@ zipimporter Objects Return the value ``__file__`` would be set to if the specified module was imported. Raise :exc:`ZipImportError` if the module couldn't be - imported. + found. .. versionadded:: 3.1 diff --git a/Lib/test/test_zipimport.py b/Lib/test/test_zipimport.py index e709c68adb206f2..3a1ed9ef64b2be8 100644 --- a/Lib/test/test_zipimport.py +++ b/Lib/test/test_zipimport.py @@ -212,6 +212,81 @@ def testPy(self): files = {TESTMOD + ".py": test_src} self.doTest(".py", files, TESTMOD) + def test_syntax_warning(self): + files = {TESTMOD + ".py": "x = 1 is 1\n"} + self.makeZip(files) + zi = zipimport.zipimporter(TEMP_ZIP) + + with warnings.catch_warnings(record=True) as caught: + warnings.simplefilter("always", SyntaxWarning) + spec = zi.find_spec(TESTMOD) + self.assertIsNotNone(spec) + self.assertEqual(caught, []) + + mod = importlib.util.module_from_spec(spec) + spec.loader.exec_module(mod) + + self.assertEqual(len(caught), 1) + self.assertIsInstance(caught[0].message, SyntaxWarning) + + def test_syntax_error_defers_to_exec(self): + files = {TESTMOD + ".py": "x =\n"} + self.makeZip(files) + zi = zipimport.zipimporter(TEMP_ZIP) + expected_path = os.path.join(TEMP_ZIP, TESTMOD + ".py") + + spec = zi.find_spec(TESTMOD) + self.assertIsNotNone(spec) + self.assertEqual(spec.origin, expected_path) + self.assertEqual(zi.get_filename(TESTMOD), expected_path) + + mod = importlib.util.module_from_spec(spec) + with self.assertRaises(SyntaxError): + spec.loader.exec_module(mod) + + def test_get_filename_no_warnings(self): + files = {TESTMOD + ".py": "x = 1 is 1\n"} + self.makeZip(files) + zi = zipimport.zipimporter(TEMP_ZIP) + + with warnings.catch_warnings(record=True) as caught: + warnings.simplefilter("always", SyntaxWarning) + zi.get_filename(TESTMOD) + + self.assertEqual(caught, []) + + def test_syntax_error_in_source_with_valid_pyc(self): + bad_src = "x =\n" + files = {TESTMOD + ".py": bad_src, + TESTMOD + pyc_ext: make_pyc(test_co, NOW, len(bad_src))} + self.makeZip(files) + zi = zipimport.zipimporter(TEMP_ZIP) + expected_path = os.path.join(TEMP_ZIP, TESTMOD + pyc_ext) + + spec = zi.find_spec(TESTMOD) + self.assertIsNotNone(spec) + self.assertEqual(spec.origin, expected_path) + + mod = importlib.util.module_from_spec(spec) + spec.loader.exec_module(mod) + self.assertEqual(mod.get_file(), expected_path) + + def test_syntax_error_in_package_init_no_fallthrough(self): + packdir = TESTPACK + os.sep + files = {packdir + "__init__.py": "x =\n", + TESTPACK + ".py": test_src} + self.makeZip(files) + zi = zipimport.zipimporter(TEMP_ZIP) + expected_path = os.path.join(TEMP_ZIP, TESTPACK, "__init__.py") + + spec = zi.find_spec(TESTPACK) + self.assertIsNotNone(spec) + self.assertEqual(spec.origin, expected_path) + + mod = importlib.util.module_from_spec(spec) + with self.assertRaises(SyntaxError): + spec.loader.exec_module(mod) + def testPyc(self): files = {TESTMOD + pyc_ext: test_pyc} self.doTest(pyc_ext, files, TESTMOD) diff --git a/Lib/zipimport.py b/Lib/zipimport.py index 37271ff375becd8..97ad77aa41ae2ae 100644 --- a/Lib/zipimport.py +++ b/Lib/zipimport.py @@ -164,11 +164,12 @@ def get_filename(self, fullname): """get_filename(fullname) -> filename string. Return the filename for the specified module or raise ZipImportError - if it couldn't be imported. + if it couldn't be found. """ # Deciding the filename requires working out where the code # would come from if the module was actually loaded - code, ispackage, modpath = _get_module_code(self, fullname) + _, _, modpath = _get_module_code( + self, fullname, compile_source=False) return modpath @@ -790,9 +791,9 @@ def _get_pyc_source(self, path): return _get_data(self.archive, toc_entry) -# Get the code object associated with the module specified by -# 'fullname'. -def _get_module_code(self, fullname): +# Get the code object associated with the module specified by 'fullname'. +# If compile_source is false, return None for source code without reading or compiling it. +def _get_module_code(self, fullname, *, compile_source=True): path = _get_module_path(self, fullname) import_error = None for suffix, isbytecode, ispackage in _zip_searchorder: @@ -804,6 +805,8 @@ def _get_module_code(self, fullname): pass else: modpath = toc_entry[0] + if not isbytecode and not compile_source: + return None, ispackage, modpath data = _get_data(self.archive, toc_entry) code = None if isbytecode: @@ -817,7 +820,6 @@ def _get_module_code(self, fullname): # bad magic number or non-matching mtime # in byte code, try next continue - modpath = toc_entry[0] return code, ispackage, modpath else: if import_error: diff --git a/Misc/NEWS.d/next/Library/2026-07-15-21-45-33.gh-issue-141276.xH7qP2.rst b/Misc/NEWS.d/next/Library/2026-07-15-21-45-33.gh-issue-141276.xH7qP2.rst new file mode 100644 index 000000000000000..c19c7bbd31eb488 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2026-07-15-21-45-33.gh-issue-141276.xH7qP2.rst @@ -0,0 +1,6 @@ +Fix duplicate :exc:`SyntaxWarning` messages when importing source modules +from ZIP archives. :meth:`zipimport.zipimporter.get_filename`, and therefore +:meth:`~zipimport.zipimporter.find_spec`, no longer compile source modules. A +:exc:`SyntaxError` in a zipped source module is now raised when the module is +executed rather than when it is found, matching the behavior of +filesystem-based loaders.