Solve InterviewCake's word-cloud problem

Write a function to count the frequency of words in a sentence. Ignore casing
for words; ignore punctuation.

2 files changed, 80 insertions, 1 deletions

diff --git a/scratch/deepmind/part_two/todo.org b/scratch/deepmind/part_two/todo.org
index a710e05b83b8..3562dbe5be55 100644
--- a/scratch/deepmind/part_two/todo.org
+++ b/scratch/deepmind/part_two/todo.org
@@ -7,7 +7,7 @@
 * Hashing and hash tables
 ** DONE Inflight Entertainment
 ** DONE Permutation Palindrome
-** TODO Word Cloud Data
+** DONE Word Cloud Data
 ** TODO Top Scores
 * Greedy Algorithms
 ** TODO Apple Stocks
diff --git a/scratch/deepmind/part_two/word-cloud.py b/scratch/deepmind/part_two/word-cloud.py
new file mode 100644
index 000000000000..36ace8405f71
--- /dev/null
+++ b/scratch/deepmind/part_two/word-cloud.py
@@ -0,0 +1,79 @@
+import unittest
+import re
+from collections import Counter
+
+
+class WordCloudData(object):
+    def __init__(self, x):
+        x = x.replace('...', ' ').replace(' - ', ' ')
+        x = ''.join(c for c in x if c not in ',.!?;:')
+        self.words_to_counts = dict(
+            Counter(x.lower() for x in re.split(r'\s+', x)))
+
+
+# Tests
+class Test(unittest.TestCase):
+    def test_simple_sentence(self):
+        input = 'I like cake'
+
+        word_cloud = WordCloudData(input)
+        actual = word_cloud.words_to_counts
+
+        expected = {'i': 1, 'like': 1, 'cake': 1}
+        self.assertEqual(actual, expected)
+
+    def test_longer_sentence(self):
+        input = 'Chocolate cake for dinner and pound cake for dessert'
+
+        word_cloud = WordCloudData(input)
+        actual = word_cloud.words_to_counts
+
+        expected = {
+            'and': 1,
+            'pound': 1,
+            'for': 2,
+            'dessert': 1,
+            'chocolate': 1,
+            'dinner': 1,
+            'cake': 2,
+        }
+        self.assertEqual(actual, expected)
+
+    def test_punctuation(self):
+        input = 'Strawberry short cake? Yum!'
+
+        word_cloud = WordCloudData(input)
+        actual = word_cloud.words_to_counts
+
+        expected = {'cake': 1, 'strawberry': 1, 'short': 1, 'yum': 1}
+        self.assertEqual(actual, expected)
+
+    def test_hyphenated_words(self):
+        input = 'Dessert - mille-feuille cake'
+
+        word_cloud = WordCloudData(input)
+        actual = word_cloud.words_to_counts
+
+        expected = {'cake': 1, 'dessert': 1, 'mille-feuille': 1}
+        self.assertEqual(actual, expected)
+
+    def test_ellipses_between_words(self):
+        input = 'Mmm...mmm...decisions...decisions'
+
+        word_cloud = WordCloudData(input)
+        actual = word_cloud.words_to_counts
+
+        expected = {'mmm': 2, 'decisions': 2}
+        self.assertEqual(actual, expected)
+
+    def test_apostrophes(self):
+        input = "Allie's Bakery: Sasha's Cakes"
+
+        word_cloud = WordCloudData(input)
+        actual = word_cloud.words_to_counts
+
+        expected = {"bakery": 1, "cakes": 1, "allie's": 1, "sasha's": 1}
+        self.assertEqual(actual, expected)
+
+
+unittest.main(verbosity=2)