jamsa-server

Blind helper project (server)

Overview

This project implements the server side of the Jamsa Project. It provides a REST API to get and store actions.

Design

An API REST has been implemented because it is very light and it allows us to separate the RFID tags from the content. So updating the "content" of a tag only requires a simple POST request to the API.

The information in the server is stored in YAML files (one per tag).

• Getting the information from a tag ID only takes one simple GET request:
  • status: 0 if OK and -1 if ERROR.
  • id: tag id (useful for the client to check if the answer matches the request).
  • command: action that the client has to execute.
  • data: arguments for the command.

curl http://server/id

where server is the server where jamsa-server is installed and id is the tag ID. The following is the request and JSON response from the server for the tag ID 1234:

curl http://server/1234
{
   "status" : 0,
   "data" : "This is a sample",
   "id" : "1234",
   "command" : "speech"
}

Therefore the client will execute the speech command, that is a text to speech action over the data field.

• Store the information of a tag.

curl -X POST -d "tag content" http://server/id/action

where id is the tag ID and action is the command associated to that tag. The following is the request and response from the server for a tag ID 1234 with the command speech:

curl -X POST -d "tag content" http://server/1234/speech
{
    "status": 0
} 

status is 0 if OK and -1 if ERROR. The following is the YAML file content in the server for the tag ID 1234.

$ cat 1234.yaml 
---
command: speech
data: tag content