twitter_wordcloud

This is a python app which gets data from the Twitter Sample Stream and returns the occurence count for every word in the stream or for a number of words specified by the user in JSON format. It can be used as the backend for a wordcloud service.

The app is enclosed in a docker container and uses docker-compose to communicate with a redis server which is running in another docker container.

To use the app, you have to register a twitter app to get a consumer key and access token for accessing the stream, and set them in wordcloud.py at lines 12-15.

How to run the app inside the docker container:

(this will download the docker images for redis and python:2.7.9)

1. Install docker and docker-compose
2. To run the app with the default args: sudo docker-compose up (this will gather data for 2 seconds from the stream and print the first 4 words ordered by occurence count, by running the command in docker-compose.yml)
3. To run the app with custom command line args, you can use
   sudo docker-compose run web ./wordcloud.py [seconds_of_streaming] [max_words]

How to build the app locally (without docker):

1. Make sure you have python version 2.7.9 installed. For older versions, use pip install requests[security] so that the requests made by urllib3 won't throw security exceptions.
2. Install redis pip install redis (python 2.7.9 comes with pip)
3. Install tweepy pip install tweepy
4. Start the redis server
5. Replace the value of the redis host at line 19 in wordcloud.py with the IP of the redis server. You can use 'localhost' if the server is local.
6. Run ./wordcloud.py [seconds_of_streaming] [max_words]