Spring Boot HL7v2 support

ipf-hl7-spring-boot-starter sets up the infrastructure for HL7v2-based IHE transactions.

The dependency on the IPF Spring Boot IHE HL7 starter module is:

    <dependency>
        <groupId>org.openehealth.ipf.boot</groupId>
        <artifactId>ipf-hl7-spring-boot-starter</artifactId>
    </dependency>

ipf-hl7-spring-boot-starter auto-configures by default:

a org.apache.camel.component.hl7.HL7MLLPCodec
a org.openehealth.ipf.modules.hl7.parser.CustomModelClassFactory covering IHE-specific PIX and PDQ message structures
a ca.uhn.hl7v2.conf.store.ProfileStore referencing predefined IHE Gazelle profiles
a ca.uhn.hl7v2.validation.ValidationContext without any specific validation rules
a configurable ca.uhn.hl7v2.parser.ParserConfiguration

Furthermore, if a single org.springframework.cache.CacheManager bean is available and the application property ipf.hl7v2.caching is set to true, the following caching storage beans are set up:

interactiveContinuationStorage for interactive continuation
unsolicitedFragmentationStorage for unsolicited fragmentation

You can define your own beans of this type in order to override the defaults.

The module instantiates a mandatory and correctly configured ca.uhn.hl7v2.HapiContext. The instance can, however, be customized by providing a HapiContextCustomizer bean:

    @Bean
    public HapiContextCustomizer hapiContextCustomizer() {
        return new HapiContextCustomizer() {
            
            public void customizeHapiContext(HapiContext fhirContext) {
                // configure HapiContext here
            }
        }
    }

The actual cache implementation being used is the one that Spring Boot finds on the classpath.

ipf-hl7-spring-boot-starter provides the following application properties:

Property (`ipf.hl7v2.`)	Default	Description
`charset`	UTF-8	Charset for HL7v2 messages
`convert-line-feed`	false	Whether to convert line feeds to proper segment separators before parsing starts
`caching`	false	Whether to set up caches for paging and unsolicited fragmentation
`generator`	file	ID generator for HL7 messages. One of “file”, “uuid”, “nano”.

In case you use the file-based ID generator, you configure it as follows:

Property (`ipf.hl7v2.id-generator`)	Default	Description
`lo`	100	How many IDs to be generated internally before incrementing the file value
`directory`	Value of `Home.getHomeDirectory().getAbsolutePath();`	Directory of the ID file
`file-name`	id_file	Name of the file
`never-fail`	true	If set to false retrieving a new ID may fail if the ID file can not be written/read. If set to true, failures will be ignored, which means that IDs may be repeated after a JVM restart.
`minimize-reads`	false	If set to true, the generator minimizes the number of disk reads by caching the last read value. This means one disk read less per X number of IDs generated, but also means that multiple instances of this generator may clobber each other’s values.

The instantiated ca.uhn.hl7v2.parser.ParserConfiguration can be immediately configured with properties starting with ipf.hl7v2.parser

Property (`ipf.hl7v2.parser`)	Default	Description
`allow-unknown-versions`	false	If the HL7 parser accepts unknown HL7 versions
… further properties of ParserConfiguration	…	…

See ipf-spring-boot-starter and ipf-atna-spring-boot-starter for additional properties.