search

aehrc / pathling

Tools that make it easier to use FHIR® and clinical terminology within data analytics, built on Apache Spark.
https://pathling.csiro.au
Apache License 2.0
91 stars 12 forks source link

Using pathling and HAPI NDJson parser in same application does not mix #1779

Closed jkiddo closed 9 months ago

jkiddo commented 10 months ago

I have a project where I use pathling (v6.4.2) as well as parsing up some FHIR ndjson files (using the FhirContext.forR4().newNDJsonParser().parseResource for doing some integrity checks). It does create some issues as there are conflicting dependencies between the use of jackson between HAPI FHIR deps and pathing:

11:54:21.953 [main] [] ERROR o.s.boot.SpringApplication - Application run failed
java.lang.NoClassDefFoundError: com/fasterxml/jackson/core/StreamReadConstraints
    at ca.uhn.fhir.parser.json.jackson.JacksonStructure.createStreamReadConstraints(JacksonStructure.java:399)
    at ca.uhn.fhir.parser.json.jackson.JacksonStructure.createObjectMapper(JacksonStructure.java:393)
    at ca.uhn.fhir.parser.json.jackson.JacksonStructure.<clinit>(JacksonStructure.java:54)
    at ca.uhn.fhir.parser.JsonParser.doParseResource(JsonParser.java:240)
    at ca.uhn.fhir.parser.BaseParser.parseResource(BaseParser.java:676)
    at ca.uhn.fhir.parser.BaseParser.parseResource(BaseParser.java:731)
    at ca.uhn.fhir.parser.BaseParser.parseResource(BaseParser.java:741)
    at ca.uhn.fhir.parser.NDJsonParser.doParseResource(NDJsonParser.java:114)
    at ca.uhn.fhir.parser.BaseParser.parseResource(BaseParser.java:676)
    at ca.uhn.fhir.parser.BaseParser.parseResource(BaseParser.java:736)
    at ca.uhn.fhir.parser.BaseParser.parseResource(BaseParser.java:654)
    at com.trifork.Main.run(Main.java:54)
    at org.springframework.boot.SpringApplication.callRunner(SpringApplication.java:765)
    at org.springframework.boot.SpringApplication.lambda$callRunners$2(SpringApplication.java:749)
    at java.base/java.util.stream.ForEachOps$ForEachOp$OfRef.accept(ForEachOps.java:183)
    at java.base/java.util.stream.SortedOps$SizedRefSortingSink.end(SortedOps.java:357)
    at java.base/java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:485)
    at java.base/java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:474)
    at java.base/java.util.stream.ForEachOps$ForEachOp.evaluateSequential(ForEachOps.java:150)
    at java.base/java.util.stream.ForEachOps$ForEachOp$OfRef.evaluateSequential(ForEachOps.java:173)
    at java.base/java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
    at java.base/java.util.stream.ReferencePipeline.forEach(ReferencePipeline.java:497)
    at org.springframework.boot.SpringApplication.callRunners(SpringApplication.java:744)
    at org.springframework.boot.SpringApplication.run(SpringApplication.java:315)
    at org.springframework.boot.SpringApplication.run(SpringApplication.java:1300)
    at org.springframework.boot.SpringApplication.run(SpringApplication.java:1289)
    at com.trifork.Main.main(Main.java:31)
Caused by: java.lang.ClassNotFoundException: com.fasterxml.jackson.core.StreamReadConstraints
    at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:581)
    at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:178)
    at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:527)
    ... 27 common frames omitted

Is it possible to get the dependencies of jackson bumped to a newer version in pathling ?

johngrimes commented 10 months ago

Hi @jkiddo, thanks for creating this issue!

Which artifact are you using? If it is the library runtime, we can probably shade it in if there is a conflict.

jkiddo commented 10 months ago

HAPI FHIR 6.10.x (transitively used by Pathling v6.4.2) uses Jackson 2.15.3, but the underlying Scala module from Pathling ain't pleased about that version as sticking to that version causes the following error:

Caused by: com.fasterxml.jackson.databind.JsonMappingException: Scala module 2.15.3 requires Jackson Databind version >= 2.15.0 and < 2.16.0 - Found jackson-databind version 2.13.4

com.fasterxml.jackson.core.StreamReadConstraints was introduced in Jackson core in 2.15 which is then used in HAPI.

jkiddo commented 10 months ago

I'm using the encoders and library-api from Pathling.

johngrimes commented 10 months ago

I think it might just be a case of switching to using the library-runtime artifact instead of the library-api.

The library runtime excludes and relocates certain dependencies to make it play nice when included in environments where there are other ambient and overlapping dependencies, e.g. Spark clusters and other Java apps.

I'll update the documentation to reflect this advice.

I got a simple app working along the lines of what you are describing:

package org.example;

import au.csiro.pathling.config.HttpClientCachingConfiguration;
import au.csiro.pathling.config.TerminologyConfiguration;
import au.csiro.pathling.encoders.FhirEncoders;
import au.csiro.pathling.library.PathlingContext;
import au.csiro.pathling.library.io.source.NdjsonSource;
import ca.uhn.fhir.context.FhirContext;
import ca.uhn.fhir.validation.FhirValidator;
import ca.uhn.fhir.validation.IValidatorModule;
import ca.uhn.fhir.validation.SingleValidationMessage;
import ca.uhn.fhir.validation.ValidationResult;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;
import org.hl7.fhir.common.hapi.validation.validator.FhirInstanceValidator;
import org.hl7.fhir.instance.model.api.IBaseResource;
import org.hl7.fhir.r4.model.Enumerations.ResourceType;

public class App {

  public static void main(final String[] args) {
    final SparkSession spark = SparkSession.builder().master("local[*]").getOrCreate();

    final HttpClientCachingConfiguration cacheConfig = HttpClientCachingConfiguration.builder()
        .overrideExpiry(2628000)
        .build();

    final TerminologyConfiguration terminologyConfig = TerminologyConfiguration.builder()
        .serverUrl("https://r4.ontoserver.csiro.au/fhir")
        .cache(cacheConfig)
        .build();

    final PathlingContext pc = PathlingContext.create(spark, terminologyConfig);

    final NdjsonSource data = pc.read()
        .ndjson("/Users/gri306/Library/CloudStorage/OneDrive-CSIRO/Data/synthea/1k_20190829/fhir");

    final Dataset<Row> result = data.extract(ResourceType.PATIENT)
        .column("name.first().given.first()", "Given name")
        .column("name.first().family", "Family name")
        .column("telecom.where(system = 'phone').value", "Phone number")
        .column(
            "reverseResolve(Condition.subject).exists(code.subsumedBy(http://snomed.info/sct|56265001))",
            "Heart disease")
        .filter(
            "(reverseResolve(Immunization.patient).vaccineCode.memberOf('https://aehrc.csiro.au/fhir/ValueSet/covid-19-vaccines') contains true).not()")
        .execute();

    result.show(10, false);

    final FhirEncoders encoders = FhirEncoders.forR4().getOrCreate();
    final IBaseResource patient = data.read(ResourceType.PATIENT).as(encoders.of("Patient"))
        .collectAsList().get(0);

    final FhirContext context = encoders.getContext();
    final FhirValidator validator = context.newValidator();
    final IValidatorModule module = new FhirInstanceValidator(context);
    validator.registerValidatorModule(module);

    final ValidationResult validationResult = validator.validateWithResult(patient);

    for (final SingleValidationMessage next : validationResult.getMessages()) {
      System.out.println(next.getLocationString() + " " + next.getMessage());
    }
  }
}

And here is the pom.xml:

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
  <modelVersion>4.0.0</modelVersion>

  <groupId>org.example</groupId>
  <artifactId>pathling-java-app</artifactId>
  <version>1.0-SNAPSHOT</version>
  <packaging>jar</packaging>

  <name>pathling-java-app</name>

  <properties>
    <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    <spark.version>3.4.2</spark.version>
    <hadoop.version>3.3.4</hadoop.version>
    <pathling.version>6.4.2</pathling.version>
    <hapi.version>6.10.0</hapi.version>
  </properties>

  <dependencies>
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-core_2.12</artifactId>
      <version>${spark.version}</version>
    </dependency>
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-sql_2.12</artifactId>
      <version>${spark.version}</version>
    </dependency>
    <dependency>
      <groupId>org.apache.spark</groupId>
      <artifactId>spark-catalyst_2.12</artifactId>
      <version>${spark.version}</version>
    </dependency>
    <dependency>
      <groupId>org.apache.hadoop</groupId>
      <artifactId>hadoop-common</artifactId>
      <version>${hadoop.version}</version>
    </dependency>
    <dependency>
      <groupId>org.apache.hadoop</groupId>
      <artifactId>hadoop-hdfs</artifactId>
      <version>${hadoop.version}</version>
    </dependency>
    <dependency>
      <groupId>au.csiro.pathling</groupId>
      <artifactId>library-runtime</artifactId>
      <version>${pathling.version}</version>
    </dependency>

    <dependency>
      <groupId>ca.uhn.hapi.fhir</groupId>
      <artifactId>hapi-fhir-base</artifactId>
      <version>${hapi.version}</version>
    </dependency>
    <dependency>
      <groupId>ca.uhn.hapi.fhir</groupId>
      <artifactId>hapi-fhir-structures-r4</artifactId>
      <version>${hapi.version}</version>
    </dependency>
    <dependency>
      <groupId>ca.uhn.hapi.fhir</groupId>
      <artifactId>hapi-fhir-validation</artifactId>
      <version>${hapi.version}</version>
    </dependency>
    <dependency>
      <groupId>ca.uhn.hapi.fhir</groupId>
      <artifactId>hapi-fhir-validation-resources-r4</artifactId>
      <version>${hapi.version}</version>
    </dependency>
  </dependencies>

  <build>
    <plugins>
      <plugin>
        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-compiler-plugin</artifactId>
        <version>3.8.1</version>
        <configuration>
          <source>11</source>
          <target>11</target>
        </configuration>
      </plugin>
    </plugins>
  </build>
</project>

Give it a try and let me know how you go.

jkiddo commented 10 months ago

I found another way. Instead of using the library-runtime and stick to the library-api (thereby not having to 'restate' the spark dependencies which transitively comes along when using the library-api) and then instead control the jackson version (<jackson-bom.version>2.15.3</jackson-bom.version>) I got it to work (add to the mix, its also a Spring Boot application) with the following below. This will work as long as HAPI FHIR is within the version range >= 2.15.0 and < 2.16.0 which for my case is OK for now.

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"
         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>

    <groupId>com.trifok</groupId>
    <artifactId>nd2parquet</artifactId>
    <version>1.0-SNAPSHOT</version>

    <parent>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-parent</artifactId>
        <version>2.7.18</version>

    </parent>

    <properties>

        <maven.compiler.source>11</maven.compiler.source>
        <maven.compiler.target>11</maven.compiler.target>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
        <jackson-bom.version>2.15.3</jackson-bom.version>

    </properties>
    <build>
        <finalName>nd2parquet</finalName>

        <plugins>
            <plugin>
                <groupId>org.springframework.boot</groupId>
                <artifactId>spring-boot-maven-plugin</artifactId>
                <executions>
                    <execution>
                        <goals>
                            <goal>repackage</goal>
                        </goals>
                    </execution>
                </executions>
            </plugin>
        </plugins>
    </build>

    <dependencies>
        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter</artifactId>
        </dependency>
        <dependency>
            <groupId>au.csiro.pathling</groupId>
            <artifactId>encoders</artifactId>
            <version>6.4.2</version>
        </dependency>
        <dependency>
            <groupId>au.csiro.pathling</groupId>
            <artifactId>library-api</artifactId>
            <version>6.4.2</version>
        </dependency>
    </dependencies>
</project>
johngrimes commented 9 months ago

Good to hear that you got it working!

The FHIR server implementation within Pathling is also a Spring Boot + HAPI + Spark application, so its POM (+ the parent POM) might provide an example if you require any further guidance.

https://github.com/aehrc/pathling/blob/main/fhir-server/pom.xml