#Sentiment Demo A REST service that delegates to a remote TCP server to handle requests. This is built with the Spring Cloud Stream Binder Servlet and the [spring-cloud-starter-stream-processor-tcp-client](https://github .com/spring-cloud-stream-app-starters/tcp/tree/master/spring-cloud-starter-stream-processor-tcp-client).
This is configured to work with a Python sentiment-analyzer that
accepts a list of tweets as JSON and returns the text along with a score called polarity.
The sentiment analyzer only cares about the text, so if it is running and this app is connected to it, you can try posting a tweet:
$curl <host>:8080/stream -H'Content-type:application/json' -d '[[{"text":"v"},{"text":"tota"},{"text":"really great!!"},
{"text":"happy hay happy"}]]
[[{"text":"v"},{"text":"tota"},{"text":"really great!!"},{"text":"happy hay happy"}]]This will return what you posted, indicating that the request was accepted.
Note that this JSON is a nested list containing a single list. A bit of impedance from using the binder servlet which does some list handling itself. In this case I need to pass the entire list as the request.
To see the result:
$ curl <host>:8080/stream/sentiments
["[{\"polarity\": 0.5, \"text\": \"v\"}, {\"polarity\": 0.5, \"text\": \"tota\"}, {\"polarity\": 0.67, \"text\": \"really great!!\"}, {\"polarity\": 0.9, \"text\": \"happy hay happy\"}]"]Not pretty, but the binder servlet makes for a simple implementation and makes it easy to reuse the tcp client processer which, it turns out, is not straightforward for this type of app. And I really want to test interoperability with the tcp client processor since eventually this will be deployed with Spring Cloud Dataflow.