Prevented makemigrations from writing in sys.path[0].

There's no reason to assume that sys.path[0] is an appropriate location
for generating code. Specifically that doesn't work with extend_sys_path
which puts the additional directories at the end of sys.path.

In order to create a new migrations module, instead of using an
arbitrary entry from sys.path, import as much as possible from the path
to the module, then create missing submodules from there.

Without this change, the tests introduced in the following commit fail,
which seems sufficient to prevent regressions for such a refactoring.

1 files changed, 18 insertions, 0 deletions

diff --git a/django/utils/module_loading.py b/django/utils/module_loading.py
index 55d89caa46..5dc3aefe75 100644
--- a/django/utils/module_loading.py
+++ b/django/utils/module_loading.py
@@ -148,3 +148,21 @@ else:
         else:
             # Exhausted the search, so the module cannot be found.
             return False
+
+
+def module_dir(module):
+    """
+    Find the name of the directory that contains a module, if possible.
+
+    Raise ValueError otherwise, e.g. for namespace packages that are split
+    over several directories.
+    """
+    # Convert to list because _NamespacePath does not support indexing on 3.3.
+    paths = list(getattr(module, '__path__', []))
+    if len(paths) == 1:
+        return paths[0]
+    else:
+        filename = getattr(module, '__file__', None)
+        if filename is not None:
+            return os.path.dirname(filename)
+    raise ValueError("Cannot determine directory containing %s" % module)