Issue #13645: pyc files now contain the size of the corresponding source

code, to avoid timestamp collisions (especially on filesystems with a low
timestamp resolution) when checking for freshness of the bytecode.

8 files changed, 117 insertions, 40 deletions

diff --git a/Lib/importlib/_bootstrap.py b/Lib/importlib/_bootstrap.py
index 520c10a..5bdd111 100644
--- a/Lib/importlib/_bootstrap.py
+++ b/Lib/importlib/_bootstrap.py
@@ -331,25 +331,40 @@ class _LoaderBasics:
         filename = self.get_filename(fullname).rpartition(path_sep)[2]
         return filename.rsplit('.', 1)[0] == '__init__'
 
-    def _bytes_from_bytecode(self, fullname, data, source_mtime):
+    def _bytes_from_bytecode(self, fullname, data, source_stats):
         """Return the marshalled bytes from bytecode, verifying the magic
-        number and timestamp along the way.
+        number, timestamp and source size along the way.
 
-        If source_mtime is None then skip the timestamp check.
+        If source_stats is None then skip the timestamp check.
 
         """
         magic = data[:4]
         raw_timestamp = data[4:8]
+        raw_size = data[8:12]
         if len(magic) != 4 or magic != imp.get_magic():
             raise ImportError("bad magic number in {}".format(fullname))
         elif len(raw_timestamp) != 4:
             raise EOFError("bad timestamp in {}".format(fullname))
-        elif source_mtime is not None:
-            if marshal._r_long(raw_timestamp) != source_mtime:
-                raise ImportError("bytecode is stale for {}".format(fullname))
+        elif len(raw_size) != 4:
+            raise EOFError("bad size in {}".format(fullname))
+        if source_stats is not None:
+            try:
+                source_mtime = int(source_stats['mtime'])
+            except KeyError:
+                pass
+            else:
+                if marshal._r_long(raw_timestamp) != source_mtime:
+                    raise ImportError("bytecode is stale for {}".format(fullname))
+            try:
+                source_size = source_stats['size'] & 0xFFFFFFFF
+            except KeyError:
+                pass
+            else:
+                if marshal._r_long(raw_size) != source_size:
+                    raise ImportError("bytecode is stale for {}".format(fullname))
         # Can't return the code object as errors from marshal loading need to
         # propagate even when source is available.
-        return data[8:]
+        return data[12:]
 
     @module_for_loader
     def _load_module(self, module, *, sourceless=False):
@@ -377,11 +392,20 @@ class SourceLoader(_LoaderBasics):
     def path_mtime(self, path):
         """Optional method that returns the modification time (an int) for the
         specified path, where path is a str.
+        """
+        raise NotImplementedError
 
-        Implementing this method allows the loader to read bytecode files.
+    def path_stats(self, path):
+        """Optional method returning a metadata dict for the specified path
+        to by the path (str).
+        Possible keys:
+        - 'mtime' (mandatory) is the numeric timestamp of last source
+          code modification;
+        - 'size' (optional) is the size in bytes of the source code.
 
+        Implementing this method allows the loader to read bytecode files.
         """
-        raise NotImplementedError
+        return {'mtime': self.path_mtime(path)}
 
     def set_data(self, path, data):
         """Optional method which writes data (bytes) to a file path (a str).
@@ -407,7 +431,7 @@ class SourceLoader(_LoaderBasics):
     def get_code(self, fullname):
         """Concrete implementation of InspectLoader.get_code.
 
-        Reading of bytecode requires path_mtime to be implemented. To write
+        Reading of bytecode requires path_stats to be implemented. To write
         bytecode, set_data must also be implemented.
 
         """
@@ -416,10 +440,11 @@ class SourceLoader(_LoaderBasics):
         source_mtime = None
         if bytecode_path is not None:
             try:
-                source_mtime = self.path_mtime(source_path)
+                st = self.path_stats(source_path)
             except NotImplementedError:
                 pass
             else:
+                source_mtime = int(st['mtime'])
                 try:
                     data = self.get_data(bytecode_path)
                 except IOError:
@@ -427,7 +452,7 @@ class SourceLoader(_LoaderBasics):
                 else:
                     try:
                         bytes_data = self._bytes_from_bytecode(fullname, data,
-                                                               source_mtime)
+                                                               st)
                     except (ImportError, EOFError):
                         pass
                     else:
@@ -448,6 +473,7 @@ class SourceLoader(_LoaderBasics):
             # throw an exception.
             data = bytearray(imp.get_magic())
             data.extend(marshal._w_long(source_mtime))
+            data.extend(marshal._w_long(len(source_bytes)))
             data.extend(marshal.dumps(code_object))
             try:
                 self.set_data(bytecode_path, data)
@@ -492,9 +518,10 @@ class _SourceFileLoader(_FileLoader, SourceLoader):
 
     """Concrete implementation of SourceLoader using the file system."""
 
-    def path_mtime(self, path):
-        """Return the modification time for the path."""
-        return int(_os.stat(path).st_mtime)
+    def path_stats(self, path):
+        """Return the metadat for the path."""
+        st = _os.stat(path)
+        return {'mtime': st.st_mtime, 'size': st.st_size}
 
     def set_data(self, path, data):
         """Write bytes data to a file."""
diff --git a/Lib/importlib/abc.py b/Lib/importlib/abc.py
index df8cd93..0d37629 100644
--- a/Lib/importlib/abc.py
+++ b/Lib/importlib/abc.py
@@ -123,7 +123,20 @@ class SourceLoader(_bootstrap.SourceLoader, ResourceLoader, ExecutionLoader):
 
     def path_mtime(self, path):
         """Return the (int) modification time for the path (str)."""
-        raise NotImplementedError
+        if self.path_stats.__func__ is SourceLoader.path_stats:
+            raise NotImplementedError
+        return int(self.path_stats(path)['mtime'])
+
+    def path_stats(self, path):
+        """Return a metadata dict for the source pointed to by the path (str).
+        Possible keys:
+        - 'mtime' (mandatory) is the numeric timestamp of last source
+          code modification;
+        - 'size' (optional) is the size in bytes of the source code.
+        """
+        if self.path_mtime.__func__ is SourceLoader.path_mtime:
+            raise NotImplementedError
+        return {'mtime': self.path_mtime(path)}
 
     def set_data(self, path, data):
         """Write the bytes to the path (if possible).
diff --git a/Lib/importlib/test/source/test_abc_loader.py b/Lib/importlib/test/source/test_abc_loader.py
index 3c19b0b..74ca2e1 100644
--- a/Lib/importlib/test/source/test_abc_loader.py
+++ b/Lib/importlib/test/source/test_abc_loader.py
@@ -5,6 +5,7 @@ from .. import abc as testing_abc
 from .. import util
 from . import util as source_util
 
+import collections
 import imp
 import inspect
 import io
@@ -40,8 +41,10 @@ class SourceLoaderMock(SourceOnlyLoaderMock):
     def __init__(self, path, magic=imp.get_magic()):
         super().__init__(path)
         self.bytecode_path = imp.cache_from_source(self.path)
+        self.source_size = len(self.source)
         data = bytearray(magic)
         data.extend(marshal._w_long(self.source_mtime))
+        data.extend(marshal._w_long(self.source_size))
         code_object = compile(self.source, self.path, 'exec',
                                 dont_inherit=True)
         data.extend(marshal.dumps(code_object))
@@ -56,9 +59,9 @@ class SourceLoaderMock(SourceOnlyLoaderMock):
         else:
             raise IOError
 
-    def path_mtime(self, path):
+    def path_stats(self, path):
         assert path == self.path
-        return self.source_mtime
+        return {'mtime': self.source_mtime, 'size': self.source_size}
 
     def set_data(self, path, data):
         self.written[path] = bytes(data)
@@ -657,6 +660,7 @@ class SourceLoaderBytecodeTests(SourceLoaderTestHarness):
             self.assertIn(self.cached, self.loader.written)
             data = bytearray(imp.get_magic())
             data.extend(marshal._w_long(self.loader.source_mtime))
+            data.extend(marshal._w_long(self.loader.source_size))
             data.extend(marshal.dumps(code_object))
             self.assertEqual(self.loader.written[self.cached], bytes(data))
 
@@ -847,7 +851,7 @@ class AbstractMethodImplTests(unittest.TestCase):
         # Required abstractmethods.
         self.raises_NotImplementedError(ins, 'get_filename', 'get_data')
         # Optional abstractmethods.
-        self.raises_NotImplementedError(ins,'path_mtime', 'set_data')
+        self.raises_NotImplementedError(ins,'path_stats', 'set_data')
 
     def test_PyLoader(self):
         self.raises_NotImplementedError(self.PyLoader(), 'source_path',
diff --git a/Lib/importlib/test/source/test_file_loader.py b/Lib/importlib/test/source/test_file_loader.py
index de1c4d8..0809077 100644
--- a/Lib/importlib/test/source/test_file_loader.py
+++ b/Lib/importlib/test/source/test_file_loader.py
@@ -70,11 +70,6 @@ class SimpleTest(unittest.TestCase):
             module_dict_id = id(module.__dict__)
             with open(mapping['_temp'], 'w') as file:
                 file.write("testing_var = 42\n")
-            # For filesystems where the mtime is only to a second granularity,
-            # everything that has happened above can be too fast;
-            # force an mtime on the source that is guaranteed to be different
-            # than the original mtime.
-            loader.path_mtime = self.fake_mtime(loader.path_mtime)
             module = loader.load_module('_temp')
             self.assertTrue('testing_var' in module.__dict__,
                          "'testing_var' not in "
@@ -190,10 +185,17 @@ class BadBytecodeTest(unittest.TestCase):
                                                 del_source=del_source)
             test('_temp', mapping, bc_path)
 
+    def _test_partial_size(self, test, *, del_source=False):
+        with source_util.create_modules('_temp') as mapping:
+            bc_path = self.manipulate_bytecode('_temp', mapping,
+                                                lambda bc: bc[:11],
+                                                del_source=del_source)
+            test('_temp', mapping, bc_path)
+
     def _test_no_marshal(self, *, del_source=False):
         with source_util.create_modules('_temp') as mapping:
             bc_path = self.manipulate_bytecode('_temp', mapping,
-                                                lambda bc: bc[:8],
+                                                lambda bc: bc[:12],
                                                 del_source=del_source)
             file_path = mapping['_temp'] if not del_source else bc_path
             with self.assertRaises(EOFError):
@@ -202,7 +204,7 @@ class BadBytecodeTest(unittest.TestCase):
     def _test_non_code_marshal(self, *, del_source=False):
         with source_util.create_modules('_temp') as mapping:
             bytecode_path = self.manipulate_bytecode('_temp', mapping,
-                                    lambda bc: bc[:8] + marshal.dumps(b'abcd'),
+                                    lambda bc: bc[:12] + marshal.dumps(b'abcd'),
                                     del_source=del_source)
             file_path = mapping['_temp'] if not del_source else bytecode_path
             with self.assertRaises(ImportError):
@@ -211,7 +213,7 @@ class BadBytecodeTest(unittest.TestCase):
     def _test_bad_marshal(self, *, del_source=False):
         with source_util.create_modules('_temp') as mapping:
             bytecode_path = self.manipulate_bytecode('_temp', mapping,
-                                                lambda bc: bc[:8] + b'<test>',
+                                                lambda bc: bc[:12] + b'<test>',
                                                 del_source=del_source)
             file_path = mapping['_temp'] if not del_source else bytecode_path
             with self.assertRaises(EOFError):
@@ -235,7 +237,7 @@ class SourceLoaderBadBytecodeTest(BadBytecodeTest):
         def test(name, mapping, bytecode_path):
             self.import_(mapping[name], name)
             with open(bytecode_path, 'rb') as file:
-                self.assertGreater(len(file.read()), 8)
+                self.assertGreater(len(file.read()), 12)
 
         self._test_empty_file(test)
 
@@ -243,7 +245,7 @@ class SourceLoaderBadBytecodeTest(BadBytecodeTest):
         def test(name, mapping, bytecode_path):
             self.import_(mapping[name], name)
             with open(bytecode_path, 'rb') as file:
-                self.assertGreater(len(file.read()), 8)
+                self.assertGreater(len(file.read()), 12)
 
         self._test_partial_magic(test)
 
@@ -254,7 +256,7 @@ class SourceLoaderBadBytecodeTest(BadBytecodeTest):
         def test(name, mapping, bytecode_path):
             self.import_(mapping[name], name)
             with open(bytecode_path, 'rb') as file:
-                self.assertGreater(len(file.read()), 8)
+                self.assertGreater(len(file.read()), 12)
 
         self._test_magic_only(test)
 
@@ -276,11 +278,22 @@ class SourceLoaderBadBytecodeTest(BadBytecodeTest):
         def test(name, mapping, bc_path):
             self.import_(mapping[name], name)
             with open(bc_path, 'rb') as file:
-                self.assertGreater(len(file.read()), 8)
+                self.assertGreater(len(file.read()), 12)
 
         self._test_partial_timestamp(test)
 
     @source_util.writes_bytecode_files
+    def test_partial_size(self):
+        # When the size is partial, regenerate the .pyc, else
+        # raise EOFError.
+        def test(name, mapping, bc_path):
+            self.import_(mapping[name], name)
+            with open(bc_path, 'rb') as file:
+                self.assertGreater(len(file.read()), 12)
+
+        self._test_partial_size(test)
+
+    @source_util.writes_bytecode_files
     def test_no_marshal(self):
         # When there is only the magic number and timestamp, raise EOFError.
         self._test_no_marshal()
@@ -375,6 +388,13 @@ class SourcelessLoaderBadBytecodeTest(BadBytecodeTest):
 
         self._test_partial_timestamp(test, del_source=True)
 
+    def test_partial_size(self):
+        def test(name, mapping, bytecode_path):
+            with self.assertRaises(EOFError):
+                self.import_(bytecode_path, name)
+
+        self._test_partial_size(test, del_source=True)
+
     def test_no_marshal(self):
         self._test_no_marshal(del_source=True)
 
diff --git a/Lib/pkgutil.py b/Lib/pkgutil.py
index bdea23e..ef027be 100644
--- a/Lib/pkgutil.py
+++ b/Lib/pkgutil.py
@@ -21,7 +21,7 @@ def read_code(stream):
     if magic != imp.get_magic():
         return None
 
-    stream.read(4) # Skip timestamp
+    stream.read(8) # Skip timestamp and size
     return marshal.load(stream)
 
 
diff --git a/Lib/py_compile.py b/Lib/py_compile.py
index 5adb70a..62d69ad 100644
--- a/Lib/py_compile.py
+++ b/Lib/py_compile.py
@@ -110,9 +110,11 @@ def compile(file, cfile=None, dfile=None, doraise=False, optimize=-1):
     """
     with tokenize.open(file) as f:
         try:
-            timestamp = int(os.fstat(f.fileno()).st_mtime)
+            st = os.fstat(f.fileno())
         except AttributeError:
-            timestamp = int(os.stat(file).st_mtime)
+            st = os.stat(file)
+        timestamp = int(st.st_mtime)
+        size = st.st_size & 0xFFFFFFFF
         codestring = f.read()
     try:
         codeobject = builtins.compile(codestring, dfile or file, 'exec',
@@ -139,6 +141,7 @@ def compile(file, cfile=None, dfile=None, doraise=False, optimize=-1):
     with open(cfile, 'wb') as fc:
         fc.write(b'\0\0\0\0')
         wr_long(fc, timestamp)
+        wr_long(fc, size)
         marshal.dump(codeobject, fc)
         fc.flush()
         fc.seek(0, 0)
diff --git a/Lib/test/test_import.py b/Lib/test/test_import.py
index 9f80b70..e4d5332 100644
--- a/Lib/test/test_import.py
+++ b/Lib/test/test_import.py
@@ -380,7 +380,7 @@ func_filename = func.__code__.co_filename
     def test_foreign_code(self):
         py_compile.compile(self.file_name)
         with open(self.compiled_name, "rb") as f:
-            header = f.read(8)
+            header = f.read(12)
             code = marshal.load(f)
         constants = list(code.co_consts)
         foreign_code = test_main.__code__
@@ -644,6 +644,16 @@ class PycacheTests(unittest.TestCase):
         self.assertEqual(sys.modules['pep3147.foo'].__cached__,
                          os.path.join(os.curdir, foo_pyc))
 
+    def test_recompute_pyc_same_second(self):
+        # Even when the source file doesn't change timestamp, a change in
+        # source size is enough to trigger recomputation of the pyc file.
+        __import__(TESTFN)
+        unload(TESTFN)
+        with open(self.source, 'a') as fp:
+            print("x = 5", file=fp)
+        m = __import__(TESTFN)
+        self.assertEqual(m.x, 5)
+
 
 class RelativeImportFromImportlibTests(test_relative_imports.RelativeImports):
 
diff --git a/Lib/test/test_zipimport.py b/Lib/test/test_zipimport.py
index 56141ef..358910b 100644
--- a/Lib/test/test_zipimport.py
+++ b/Lib/test/test_zipimport.py
@@ -19,7 +19,7 @@ import io
 from traceback import extract_tb, extract_stack, print_tb
 raise_src = 'def do_raise(): raise TypeError\n'
 
-def make_pyc(co, mtime):
+def make_pyc(co, mtime, size):
     data = marshal.dumps(co)
     if type(mtime) is type(0.0):
         # Mac mtimes need a bit of special casing
@@ -27,14 +27,14 @@ def make_pyc(co, mtime):
             mtime = int(mtime)
         else:
             mtime = int(-0x100000000 + int(mtime))
-    pyc = imp.get_magic() + struct.pack("<i", int(mtime)) + data
+    pyc = imp.get_magic() + struct.pack("<ii", int(mtime), size & 0xFFFFFFFF) + data
     return pyc
 
 def module_path_to_dotted_name(path):
     return path.replace(os.sep, '.')
 
 NOW = time.time()
-test_pyc = make_pyc(test_co, NOW)
+test_pyc = make_pyc(test_co, NOW, len(test_src))
 
 
 TESTMOD = "ziptestmodule"
@@ -293,7 +293,7 @@ class UncompressedZipImportTestCase(ImportHooksBaseTestCase):
             return __file__
         if __loader__.get_data("some.data") != b"some data":
             raise AssertionError("bad data")\n"""
-        pyc = make_pyc(compile(src, "<???>", "exec"), NOW)
+        pyc = make_pyc(compile(src, "<???>", "exec"), NOW, len(src))
         files = {TESTMOD + pyc_ext: (NOW, pyc),
                  "some.data": (NOW, "some data")}
         self.doTest(pyc_ext, files, TESTMOD)
@@ -313,7 +313,7 @@ class UncompressedZipImportTestCase(ImportHooksBaseTestCase):
         self.doTest(".py", files, TESTMOD, call=self.assertModuleSource)
 
     def testGetCompiledSource(self):
-        pyc = make_pyc(compile(test_src, "<???>", "exec"), NOW)
+        pyc = make_pyc(compile(test_src, "<???>", "exec"), NOW, len(test_src))
         files = {TESTMOD + ".py": (NOW, test_src),
                  TESTMOD + pyc_ext: (NOW, pyc)}
         self.doTest(pyc_ext, files, TESTMOD, call=self.assertModuleSource)