Implement type extraciton from data, for Dimension class

cr/cube/dimension.py

@@ -0,0 +1,61 @@
+# coding: utf-8
+
+MULTIPLE_RESPONSE_TYPE = -127
+CATEGORICAL = 'categorical'
+ENUM = 'enum'
+SELECTION_TYPE_IDS = (1, 0, -1)
+
+
+class Dimension(object):
+
+    @classmethod
+    def _contains_type(cls, elements, type_):
+        return any(type_ == element['id'] for element in elements)
+
+    @classmethod
+    def _is_uniform(cls, data):
+        return bool(data['references'].get('uniform_basis'))
+
+    @classmethod
+    def _is_selections(cls, data):
+        data_type = data['type']
+
+        if data_type['class'] != CATEGORICAL:
+            return False
+
+        category_ids = tuple(
+            category['id'] for category in data_type['categories']
+        )
+        if category_ids == SELECTION_TYPE_IDS and not cls._is_uniform(data):
+            return True
+
+        return False
+
+    @classmethod
+    def _is_categorical(cls, data):
+        data_class = data['type']['class']
+        return (
+            data_class == CATEGORICAL and
+            not cls._is_selections(data)
+        )
+
+    @classmethod
+    def _is_multiple_response(cls, data):
+        data_type = data['type']
+        data_class = data_type['class']
+        return (
+            data_class == ENUM and
+            cls._contains_type(data_type['elements'], MULTIPLE_RESPONSE_TYPE)
+        )
+
+    @classmethod
+    def _get_data_type(cls, data):
+        if cls._is_selections(data):
+            return 'selections'
+        elif cls._is_categorical(data):
+            return 'categorical'
+        elif cls._is_multiple_response(data):
+            return 'multiple_response'
+        elif data['type'].get('subtype'):
+            return data['type']['subtype']['class']
+        raise ValueError('Could not extract data type from: {}'.format(data))