Creating your first topology

Now, we'll create a Storm topology that breaks sentences into words and then counts the number of occurrences of each word. Implementing this topology in Storm requires the following components:

• Sentence spout (randomsentence.py): A topology always begins with a spout; that's how data gets into Storm. The sentence spout will emit an infinite stream of sentences.

• Splitter bolt (splitsentence.py): This receives sentences and splits them into words.

• Word count bolt (wordcount.py): This receives words and counts the occurrences. For each word processed, output the word along with the number of occurrences.

The following figure shows how data flows through the topology:

Word count topology

Now that we've seen the basic data flow, let's implement the topology and see how it works.

Sentence spout

In this section, we implement a spout that generates random sentences. Enter this code in a file called randomsentence.py:

import time
import random

from petrel import storm
from petrel...