Support iterating over TdmsFile and TdmsGroup

Because these act like dictionaries, iterating over them
produces group and channel names.

nptdms/tdms.py

@@ -172,6 +172,9 @@ def close(self):
             self._reader.close()
             self._reader = None
 
+    def __iter__(self):
+        return iter(self._groups)
+
     def __getitem__(self, group_name):
         """ Retrieve a TDMS group from the file by name
         """
@@ -426,6 +429,9 @@ def as_dataframe(self, time_index=False, absolute_time=False, scaled_data=True):
 
         return pandas_export.from_group(self, time_index, absolute_time, scaled_data)
 
+    def __iter__(self):
+        return iter(self._channels)
+
     def __getitem__(self, channel_name):
         """ Retrieve a TDMS channel from this group by name
         """

nptdms/test/test_tdms_file.py

@@ -218,7 +218,22 @@ def test_iterate_channel_data_in_read_mode():
             compare_arrays(actual_data, expected_channel_data)
 
 
-def test_invalid_offset_throws():
+def test_iterate_file_and_groups():
+    """ Test iterating over TdmsFile and TdmsGroup uses key values
+    """
+    test_file, expected_data = scenarios.chunked_segment().values
+
+    with test_file.get_tempfile() as temp_file:
+        tdms_file = TdmsFile.read(temp_file.file)
+        for group_name in tdms_file:
+            group = tdms_file[group_name]
+            for channel_name in group:
+                channel = group[channel_name]
+                expected_channel_data = expected_data[(group_name, channel_name)]
+                compare_arrays(channel.data, expected_channel_data)
+
+
+def test_invalid_offset_in_read_data_throws():
     """ Exception is thrown when reading a subset of data with an invalid offset
     """
     test_file, expected_data = scenarios.single_segment_with_one_channel().values
@@ -230,7 +245,7 @@ def test_invalid_offset_throws():
             assert "offset must be non-negative" in str(exc_info.value)
 
 
-def test_invalid_length_throws():
+def test_invalid_length_in_read_data_throws():
     """ Exception is thrown when reading a subset of data with an invalid length
     """
     test_file, expected_data = scenarios.single_segment_with_one_channel().values