- Architecture of the solution
- Process API Call
- Registry API Call
- Some example calls to the blood pressure calculator
- Note on compiling the server with Maven
- Next steps
The Open mHealth engine has been developed with the support of DPU, following the Scenario 1
The main classes are described in the following picture
- the DataProcessUnitController is the access point to the DPUs. It exposes a REST interface with the two DPU methods. Given the of the requests dynamically looks for DPUs and forwards the JSON call to them. It also logs requests and responses for better tracing and debugging. It catches the exceptions thrown by the engine and converts them into the proper error message and HTTP code.
- the DataProcessUnit interface declares the methods that all the DPUs have to implement.
- the DataProcessUnitBaseImpl is a base implementation for all the DPUs. It doesn't implement the DataProcessUnit interface but some utility methods used the DPUs.
- the BloodPressureDPU is a DPU implementation that calculates the (Blood Pressure Category)[http://www.heart.org/HEARTORG/Conditions/HighBloodPressure/AboutHighBloodPressure/Understanding-Blood-Pressure-Readings_UCM_301764_Article.jsp], given the Systolic and Distolic blood pressure. It is automatically registered to the DataProcessUnitController through a naming convention:
- every DPU must be in a subpackage of the DataProcessUnit package
- every DPU must be called + "DPU", where is the name of the process the DPU implements and that will be used in the REST service endpoint.
- BusinessException and SystemException are the exceptions thrown by the whole engine. They have an associated error dictionary with the error code and relative messages.
This is the main worker call into a DPU. Process receives data that belongs to one schema ID-version pair and returns data that belongs to another schema ID-version pair. There are cases where the input and output schema ID-version pairs could be the same.
The process endpoint must include support for a parameter that instructs it to preserve the original raw input data. (in the current version is not implemented)
The request URL has one path.
- The path to send some data to be processed is:
dpu/{processName}.- This will return the result of the application of
{processName}to the input data.
- This will return the result of the application of
- The HTTP method is a POST
- The payload follows the schema id and version described by the readRegistry operation
In case of success, the response is a JSON object following the schema id and version described by the readRegistry operation. In case of error, the response has a 500 status code and is a JSON object following the omh:dpu:SystemException and omh:dpu:BusinessException schema.
Registry functions much like the DSU registry. It returns the schema ID-version pairs that can be processed and the schema-ID version pairs that are returned.
The request URL has one path.
- The path to send some data to be processed is:
dpu/{processName}.- This will return the result of the application of
{processName}to the input data.
- This will return the result of the application of
- The HTTP method is a GET
- The result is the schema ID-version pairs that can be processed and the schema-ID version pairs that are returned.
In case of success, the response is a JSON object containing the schema ID-version pairs that can be processed and the schema-ID version pairs that are returned by the corresponding process method. In case of error, the response has a 500 status code and is a JSON object following the omh:dpu:SystemException and omh:dpu:BusinessException schema.
A DPU that calculates the blood pressure categories as described in an example has been developed.
The following are some example calls along with the result produced by the DPU engine
Right request:
curl -v http://localhost:8080/omh-engine-1.0.0/dpu/bloodPressure -X POST -H "Content-Type: application/json"
-d '{"systolic":100, "diastolic":99}' ; echo ""
Example HTTP request header
POST /omh-engine-1.0.0/dpu/bloodPressure HTTP/1.1
User-Agent: curl/7.30.0
Host: localhost:8080
Accept: */*
Content-Type: application/json
Content-Length: 32
Example HTTP response header (from a tomcat servlet engine)
HTTP/1.1 200 OK
Server Apache-Coyote/1.1 is not blacklisted
Server: Apache-Coyote/1.1
Content-Type: text/plain;charset=ISO-8859-1
Content-Length: 41
Example HTTP response content
{"category":"high_blood_pressure_stage1"}
Wrong request (triggering a json not well formed error):
curl -v http://localhost:8080/omh-engine-1.0.0/dpu/bloodPressure -X POST -H "Content-Type: application/json"
-d '{"systolic":100, "diastolic":99' ; echo ""
Example HTTP response content
{"BusinessException":
{"code":"json not well formed"},
{"message":"the json input is not well formed: {"systolic":100, "diastolic":99"}}
curl -v http://localhost:8080/omh-engine-1.0.0/dpu/bloodPressure ; echo ""
Example HTTP request header
GET /omh-engine-1.0.0/dpu/bloodPressure HTTP/1.1
User-Agent: curl/7.30.0
Host: localhost:8080
Accept: */*
Example HTTP response header (from a tomcat servlet engine)
HTTP/1.1 200 OK
Server Apache-Coyote/1.1 is not blacklisted
Server: Apache-Coyote/1.1
Content-Type: application/json;charset=UTF-8
Example HTTP response content
[{"schemaId":"omh:dpu:bloodpressure",
"schemaVersion":"1"},
{"schemaId":"omh:dpu:bloodpressure",
"schemaVersion":"1"}]
The project has a maven pom that lists the library dependencies.
###Using Lombok The lombok library is used because it saves to write a lot of boilerplate code. The following instructions must be followed to install it in the IDe of choice.
If in the next steps the Concordia library will be used, in must be downloaded and compiled and installed in the local maven repository
mvn install:install-file -DgroupId=name.jenkins.paul.john -DartifactId=concordia -Dversion=1.2.4 -Dpackaging=jar -DgeneratePom=true -Dfile=
- BaseException: a possible extension could be to customize the status, depending on the error code