Started work on Census class for #4

ety/__init__.py

@@ -10,6 +10,7 @@
 
 from . import data
 from .tree import EtyTree
+from .census import Census
 from .word import Word, Language  # noqa: F401
 
 
@@ -62,3 +63,7 @@ def random_word(lang='eng'):
     row = list(filter(lambda entry: entry['a_lang'] == lang, data.etyms))
     word = choice(row)['a_word']
     return Word(word, lang)
+
+
+def census(words):
+    return Census(words)

ety/census.py

@@ -0,0 +1,30 @@
+import re
+
+from .word import Word
+
+string_types = ("".__class__, u"".__class__)
+
+
+class Census(object):
+    def __init__(self, words, lang='eng'):
+        if isinstance(words, ("".__class__, u"".__class__)):
+            words = re.split('\s+', words)
+        elif isinstance(words, (list, tuple)):
+            words = list(words)
+        else:
+            raise ValueError('words argument must be either string or list')
+
+        assert type(words) is list
+
+        self.words = []
+        for word in list(words):
+            if isinstance(word, string_types):
+                self.words.append(Word(word, lang))
+            elif isinstance(word, Word):
+                self.words.append(word)
+            else:
+                raise ValueError("Invalid word type: '%s'.\ Words must\
+                    be ety.Word objects or strings" % str(type(word)))
+
+    def __eq__(self, other):
+        return self.words == other.words

tests.py

@@ -35,6 +35,14 @@ def test_tree(self):
         self.assertGreaterEqual(len(str(
             ety.tree('fabric')).split('\n')), 4)
 
+    def test_census_words(self):
+        a = ety.census(['alphabet', 'avocado', 'guitar'])
+        b = ety.census('alphabet avocado guitar')
+        self.assertEqual(a, b)
+        self.assertTrue(ety.Word('avocado') in a.words)
+        with self.assertRaises(ValueError):
+            ety.census(['valid', ety.Word('stillvalid'), 12345])
+
 
 if __name__ == '__main__':
     unittest.main()