Serverless computing in Azure with .NET

Constructing the payload

As discussed in the preceding section, the text sentiment analytics API expects the payload to be broken down into a potentially large number of relatively small "documents". When working with short snippets of text, such as tweets or SMS messages, the entire message can be submitted for scoring as a single document. As we move to work with larger texts, the text needs to be split up to avoid reaching the 10 KB limit of a single document. As we move to work with even larger texts, the total input limit of 1 MB per API call also needs to be taken into consideration.

In our implementation of constructing the payload, we will implement a one-size-fits-all approach, which will allow us to process text of any size.

We ...