Support nested tuple unpacking in for loops

Fixed issue where unpacking nested tuples in a for loop using would raise a
"couldn't apply loop context" error if the loop context was used. The regex
used to match the for loop expression now allows the list of loop variables
to contain parenthesized sub-tuples. Pull request courtesy Matt Trescott.

For example:
~~~
for (key1, val1), (key2, val2) in itertools.pairwise(dict.items()):
    ...
~~~

This is really just "kicking the can down the road" so to speak, because
it doesn't allow an infinite number of layers of tuples, but
it helps somewhat.

Closes: #368
Pull-request: https://github.com/sqlalchemy/mako/pull/368
Pull-request-sha: 3f15a87266a36306826d460cddc7699dd62a9c43

Change-Id: I52915acb8904daf7071d8c92e1de352f200131ec

1 files changed, 10 insertions, 0 deletions

diff --git a/test/test_loop.py b/test/test_loop.py
index d048ed0..2c11000 100644
--- a/test/test_loop.py
+++ b/test/test_loop.py
@@ -31,6 +31,16 @@ class TestLoop(unittest.TestCase):
                 "x",
                 "[y+1 for y in [1, 2, 3]]",
             ),
+            (
+                "for ((key1, val1), (key2, val2)) in pairwise(dict.items()):",
+                "((key1, val1), (key2, val2))",
+                "pairwise(dict.items())",
+            ),
+            (
+                "for (key1, val1), (key2, val2) in pairwise(dict.items()):",
+                "(key1, val1), (key2, val2)",
+                "pairwise(dict.items())",
+            ),
         ):
             match = _FOR_LOOP.match(statement)
             assert match and match.groups() == (target_list, expression_list)