Wondering about the mechanics of running the user's script.

1 files changed, 6 insertions, 2 deletions

diff --git a/coverage/cmdline.py b/coverage/cmdline.py
index 58eae71..0561f32 100644
--- a/coverage/cmdline.py
+++ b/coverage/cmdline.py
@@ -112,14 +112,18 @@ class CoverageScript:
 
         if settings.get('erase'):
             self.coverage.erase()
+        
         if settings.get('execute'):
             if not args:
                 help_fn("Nothing to do.")
+            # Create the runtime environment the script on the cmdline expects.
             sys.argv = args
-            self.coverage.start()
-            import __main__
             sys.path[0] = os.path.dirname(sys.argv[0])
+            import __main__ # TODO: I think this is useless...
+            self.coverage.start()
             execfile(sys.argv[0], __main__.__dict__)
+            self.coverage.stop()
+        
         if settings.get('combine'):
             self.coverage.combine()