This post is a second part of my introduction into HAPI library.

Parse HL7 message

We will start with a simple project which will demonstrate how to parse HL7 messages using HAPI API.

Tools and libraries will use

  • JDK 1.8
  • Maven
  • Eclipse (STS 3.8.1)
  • HAPI API v2.2

HAPI has a very good set of examples which is a good starting point. Will use it as well - will jump right to section ‘Parsing Messages’.

Configuration steps

  • Create empty Maven based project simple-hl7-parser using basic archetype;
  • Add HL7 dependencies: 
    ..........................
<properties>
      <hapi.version>2.2</hapi.version>
  </properties>
<dependency>
  <groupId>ca.uhn.hapi</groupId>
  <artifactId>hapi-base</artifactId>
  <version>${hapi.version}</version>
  <scope>compile</scope>
</dependency>
<dependency>
  <groupId>ca.uhn.hapi</groupId>
  <artifactId>hapi-structures-v24</artifactId>
  <version>${hapi.version}</version>
  <scope>compile</scope>
</dependency>
..........................
  • Import project into the IDE.

Notes: HAPI library packs structures for different HL7 versions in different artifacts. In my case I will use/parse HL7v2.4 messages - I only included ‘hapi-structures-v24’ artifact.

Coding

  • Create an instance HapiContext
  • Create an instance of CanonicalModelClassFactory
  • Add/set CanonicalModelClassFactory to HapiContext
  • Get parser (PipeParser)
  • Parse message

Below is a piece of sample code:

HapiContext context = new DefaultHapiContext();

CanonicalModelClassFactory mcf = new CanonicalModelClassFactory("2.4");

context.setModelClassFactory(mcf);

PipeParser parser = context.getPipeParser();

ca.uhn.hl7v2.model.v24.message.ORU_R01 msg = (ca.uhn.hl7v2.model.v24.message.ORU_R01) parser.parse(message);

The following things immediately pops to our eyes:

  • CanonicalModelClassFactory takes Hl7v2 specific version (2.4 in our case)
  • Parsing result is an instance of ORU_R01 class from package ca.uhn.hl7v2.model.v24.message

Items above means that we need to know which HL7v2 message we expect to parse and each version has its own strict model.